national survey of fishing, hunting, and wildlife ... · 12 years and older personal inter-view, 12...

National Survey of Fishing,

Hunting, and Wildlife-Associated Recreation, 2016

Design and Methodology Technical Report Issued August 2019

Acknowledgements 2016 FHWAR Design and Methodology Technical Report

P a g e | ii

DISCLAIMER

This report is released to inform interested parties of research and to encourage discussion.

Any views expressed in statistical, methodological, technical, or operational issues are those of

the authors and are not necessarily those of the U.S. Census Bureau.

ACKNOWLEDGEMENTS

The 2016 National Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR)

Design and Methodology Technical Report is a modified version of the 2016 National Survey of

Fishing, Hunting, and Wildlife-Associated Recreation Technical Report that was prepared for

internal use.

The internal Technical Report was developed under the direction of Lucinda P. Dalzell, FHWAR

Survey Director, Associate Director for Demographic Programs-Survey Operations (ADDP-SO).

Elke McLaren, FHWAR Team Lead, provided overall management and coordination. David

Hornick and Nghiep Huynh developed the chapters describing the survey and sample design,

response rates, weighting and estimation and variance estimation. Scott Novell also

contributed to the weighting and estimation documentation and documented the editing and

cleaning procedures. Tanner Cummings and Melissa Kresin contributed to the Design and

Methodology Technical Report. The FHWAR Survey team is under the direction of Eloise K.

Parker, Assistant Director for Demographic Programs-Survey Operations and Victoria Velkoff,

Associate Director for Demographic Programs.

Table of Contents 2016 FHWAR Design and Methodology Technical Report

P a g e | iii

TABLE OF CONTENTS

Chapter 1. BACKGROUND .................................................................................................. 1

Chapter 2. HISTORY OF THE NATIONAL SURVEY OF FISHING, HUNTING, AND WILDLIFE

ASSOCIATED RECREATION .......................................................................................... 2

INTRODUCTION ......................................................................................................................................... 2

MAJOR CHANGES IN THE SURVEY ............................................................................................................. 3

1955 to 1970 Surveys ............................................................................................................................ 3

1975 Survey ........................................................................................................................................... 4

1980 to 1985 Surveys ............................................................................................................................ 4

Overview of 1991 to 2016 Surveys Significant Methodological Differences ........................................ 5

1991 Survey ........................................................................................................................................... 6

1996 Survey ........................................................................................................................................... 6

2001 Survey ........................................................................................................................................... 7

2006 Survey ........................................................................................................................................... 8

2011 Survey ........................................................................................................................................... 9

2016 Survey ......................................................................................................................................... 10

Chapter 3. DESIGN OF THE NATIONAL SURVEY OF FISHING, HUNTING, AND WILDLIFE

ASSOCIATED RECREATION SURVEY SAMPLE .............................................................. 11

INTRODUCTION ....................................................................................................................................... 11

THE MASTER ADDRESS FILE (MAF) ......................................................................................................... 12

Creating and Updating The Demographic Sampling Frames .............................................................. 12

ADMINISTRATIVE DATA .......................................................................................................................... 13

SAMPLE SIZE ............................................................................................................................................ 13

2013 FHWAR Pre-Screener Test .......................................................................................................... 13

National Pre-Screener Sample Size ..................................................................................................... 14

Four States’ Sample Size Calculations ................................................................................................. 15

Census Divisions’ Sample Size Calculations ........................................................................................ 16

FORMATION AND SAMPLE SELECTION OF PSUs ..................................................................................... 17

Sample Areas....................................................................................................................................... 17

Sample Area Preparation .................................................................................................................... 18

Defining the PSUs ................................................................................................................................ 19

Self-Representing (SR) PSUs ................................................................................................................ 19


P a g e | iv

Non-Self-Representing (NSR) PSUs and Formation of Strata ............................................................. 21

Criteria and Priorities .......................................................................................................................... 22

PSU Sample Selection ......................................................................................................................... 23

Sample Design Summary Tables ......................................................................................................... 23

PRE-SCREENER SAMPLE: SAMPLE SELECTION OF HOUSEHOLDS WITHIN PSUs ..................................... 24

Subsampling of Pre-Screener Sample ................................................................................................. 25

Chapter 4. DATA COLLECTION METHODS .......................................................................... 27

INTRODUCTION ....................................................................................................................................... 27

PRE-SCREENER INTERNET/PAPER QUIESTIONNAIRE .............................................................................. 27

WAVE 1 SCREENER AND FIRST DETAIL INTERVIEW ................................................................................ 28

WAVE 2 CATI DETAIL INTERVIEW AND WAVE 2 CAPI SCREENER AND DETAIL INTERVIEW .................... 29

Wave 2 CATI Detail Interview ............................................................................................................. 29

Wave 2 CAPI Screener Interview ........................................................................................................ 30

WAVE 3 CATI AND CAPI DETAIL INTERVIEW ........................................................................................... 30

SUMMARY OF COMPLETE AND SUFFICIENT PARTIALS BY MODE OF INTERVIEW AND NUMBER OF

CONTACTS ............................................................................................................................................... 31

RESPONSE RATES .................................................................................................................................... 35

AAPOR Response Rate Calculator ....................................................................................................... 36

National Screener Response Rates ..................................................................................................... 36

Final Wave 3 Detail Person Interview Response Rates ....................................................................... 37

Chapter 5. DATA PREPARATION AND PROCESSING ........................................................... 39

INTRODUCTION ....................................................................................................................................... 39

PREPARATION TO CREATE LABEL FILES AND INPUT FILES ...................................................................... 39

Address Standardization ..................................................................................................................... 39

Telephone Research ............................................................................................................................ 39

Input Files ............................................................................................................................................ 39

DATA PROCESSING PER WAVE OF INTERVIEWING ................................................................................. 40

Pre-screener Interview Data Processing ............................................................................................. 40

Screener Interview Data Processing ................................................................................................... 40

Detail Interview Data Processing ........................................................................................................ 41

Shifting Responses .............................................................................................................................. 41

Recoding “Other Specify” for Species ................................................................................................. 42

Filling Skipped Questions .................................................................................................................... 42


P a g e | v

Backfilling Day Values ......................................................................................................................... 42

Renaming Variables ............................................................................................................................ 42

Blanking Don’t Know/Refused Responses .......................................................................................... 43

DATA PROCESSING AFTER THREE WAVES OF INTERVIEWING ................................................................ 43

Renaming Variables ............................................................................................................................ 43

Converting Percentages ...................................................................................................................... 44

Recoding 1/2 (Yes/No) to 1/0 ............................................................................................................. 44

EDITING AND IMPUTATION .................................................................................................................... 44

CREATING FINAL DATA PRODUCTS ......................................................................................................... 45

Chapter 6. WEIGHTING AND ESTIMATION ........................................................................ 47

INTRODUCTION ....................................................................................................................................... 47

SCREENER WEIGHTS ................................................................................................................................ 48

Base Weights ....................................................................................................................................... 48

Pre-screener Subsampling Factor ....................................................................................................... 48

Adjustment for Nonresponse.............................................................................................................. 48

First-Stage Ratio Adjustment .............................................................................................................. 49

Computing First-Stage Ratio Adjustment Factors ............................................................................... 50

Second-Stage Ratio Adjustment ......................................................................................................... 50

DETAIL PERSON WEIGHTS ....................................................................................................................... 51

Sportspersons Sample ......................................................................................................................... 51

Wildlife-Watching Sample .................................................................................................................. 53

Chapter 7. VARIANCE ESTIMATION................................................................................... 56

INTRODUCTION ....................................................................................................................................... 56

REPLICATE FACTORS AND WEIGHTS ....................................................................................................... 57

Step 1: Construct the Hadamard Matrix ............................................................................................. 57

Step 2: Assign Row Values .................................................................................................................. 58

Step 3: Calculation of the Replicate Factors for the FHWAR Survey .................................................. 59

Step 4: Calculation of the Replicate Weights for the FHWAR Survey ................................................. 60

Step 5: Perform the Weighting Process .............................................................................................. 60

Example of the Replication Method ................................................................................................... 60

VARIANCE ESTIMATES ............................................................................................................................. 64

Example of Replicate Variance Estimation ......................................................................................... 65


P a g e | vi

GENERALIZED VARIANCE FUNCTIONS (GVF) ........................................................................................... 66

ESTIMATING STANDARD ERRORS ........................................................................................................... 67

Example of Estimating Standard Errors and Confidence Intervals ..................................................... 68

GLOSSARY AND DESCRIPTIONS OF ACRONYMS .................................................................... 70

APPENDIX A. 2016 FHWAR MATERIALS ................................................................................ 72

ADVANCE LETTERS .................................................................................................................................. 72

FREQUENTLY ASKED QUESTIONS ............................................................................................................ 78

QUICK FACTS ........................................................................................................................................... 79

REFERENCE AID ....................................................................................................................................... 80

APPENDIX B. 2016 QUESTIONNAIRES AND SURVEY QUESTIONS ........................................... 88

PRE-SCREENER QUESTIONNAIRE ............................................................................................................ 88

ABDRIGED LIST OF WAVES 1 AND 2 SCREENER INTERVIEW QUESTIONS ............................................... 92

ABDRIGED LIST OF WAVE 3 DETAIL INTERVIEW QUESTIONS ................................................................. 95

APPENDIX C. VARIANCE ITEMS .......................................................................................... 127

SUMMARY OF SCREENER VARIABLES INCLUDED ON ALL PUBLIC USE FILES AND VARIANCE ITEMS BY

SAMPLE ................................................................................................................................................. 127

APPENDIX D. COMPARABILITY ACROSS FHWAR SURVEYS .................................................. 136

APPENDIX E. OUTCOME CODES .......................................................................................... 141

Background 2016 FHWAR Design and Methodology Technical Report

P a g e | 1

Chapter 1. BACKGROUND

The National Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR) is

designed to provide information on participation and expenditures related to fishing, hunting,

and other wildlife-related activities. The FHWAR has been conducted since 1955 and the 2016

FHWAR is the thirteenth National Survey. The 2016 FHWAR is the sixth National Survey that the

U.S. Census Bureau has conducted under an Interagency Agreement (IAA) with the U.S.

Department of the Interior, U.S. Fish and Wildlife Service (FWS). The survey is sponsored by the

Association of Fish and Wildlife Agencies (AFWA) and funded through the Multistate

Conservation Grant Programs, which is authorized by the Wildlife and Sport Fish Restoration

Programs Improvement Act of 2000.

The objective of the 2016 FHWAR was to conduct a national-level survey using methods and

techniques similar to those for previous National Surveys, from 1991 through 2011. The project

was to provide national-level results and state results for Maine, Minnesota, Oklahoma, and

Virginia that were directly comparable with the results from earlier National Surveys.

The 2016 FHWAR was administered by the Census Bureau using a multistage probability sample

that represented all 50 states and Washington, D.C. The target population for the 2016 FHWAR

was the household population. The sampling frame consisted of all valid housing units in the

July 2015 Master Address File (MAF).

After sampling, the survey was conducted in three phases: a pre-screener sample, an initial

screening of households to identify likely sportspersons and wildlife-watching participants, and

a series of follow-up interviews of selected household members to collect detailed data about

their hunting, fishing, and wildlife-related recreation during 2016. The FHWAR collects data for

a calendar year. In order to reduce recall bias, the data were collected about every four-eight

months.

The pre-screener interview consisted of a web-based instrument with a supplemental paper

questionnaire, if a web response was not received from the sample address. The screener and

detailed phase interviews were conducted in two modes: Computer-Assisted Telephone

Interviewing (CATI) and Computer-Assisted Personal Interviewing (CAPI).

The Census Bureau was responsible for processing and disseminating the data, which included

editing and imputing certain data items, creating weights and variance estimates, and

developing public use microdata files and a national report.

History 2016 FHWAR Design and Methodology Technical Report

P a g e | 2

Chapter 2. HISTORY OF THE NATIONAL SURVEY OF FISHING, HUNTING, AND WILDLIFE ASSOCIATED RECREATION

INTRODUCTION

The 2016 National Survey of FHWAR was designed to continue the data collection of the 1955

to 2011 Surveys. While complete comparability between any two Surveys cannot be achieved,

this section compares major findings of all the Surveys. The principal characteristics of the 1955

to 2016 Surveys are summarized in Table 2.1, which shows the scope and design of all 13

Surveys.

Table 2.1 Major Characteristics of Surveys: 1955 to 2016

Characteristic 1955 1960 1965 1970 1975 1980 1985 1991 1996 2001 2006 2011 2016

Survey design: Prescreening

interview mode and population of interest…………

Screening interview mode and population of interest…………

Detailed interview mode and population of interest………..

Respondent’s recall period…………

Sample sizes: Prescreening phase

(households)….. Screening phase

(households)……

Detailed phase (individuals): Fishing and hunting Wildlife watching3

Response rates: Screening phase…..

Detailed phase:

X

Com-bined

with detailed

phase

Personal inter-view,

12 years and

older

1 year

X

20,000

9,328 X

NA

X


12 years and

older


12 years and

older. Sub-

stantial partici-pants1

1 year

X

18,000

10,300 X

NA

X


9 years and

older


12 years and

older. Sub-


1 year

X

16,000

6,400 X

NA

X

Tele-phone inter-view,

6 years and

older


12 years and

older. Sub-


1 year

X

24,000

8,700 X

NA

X

Tele- phone/

personal inter-view,

6 years and older

Mail ques-

tionnaire, 9 years

and older

1 year

X

106,294

20,211 X

95

percent

X

Tele- phone/


6 years and

older


16 years and

older

1 year

X

116,025

30,291 5,997

95

percent

X

Tele- phone/


6 years and

older


16 years and

older

1 year

X

102,694

28,011 26,671

93

percent

X

Tele- phone/


6 years and

older

Tele-phone/


16 years and

older

1 year

X

102,804

23,179 22.723

95

percent

X

Tele- phone/


6 years and

older

Tele-phone/


16 years and

older

4-8

months

X

44,000

13,222 9,802

71

percent

X

Tele- phone/


6 years and

older

Tele-phone/


16 years and

older

4-8

months

X

52,508

25,070 15,303

75

percent

X

Tele- phone/


6 years and

older

Tele-phone/


16 years and

older

4-8

months

X

66,688

21,938 11,279

90

percent

X

Tele- phone/


6 years and

older

Tele-phone/


16 years and

older

4-12

months

X

30,400

11,330

9,329

77 percent

Web/ paper, 6 years and older

Tele- phone/


6 years and

older

Tele-phone/


16 years and

older

4-12

months

22,725

8,030

5,640 6,079

83

percent


P a g e | 3

Characteristic 1955 1960 1965 1970 1975 1980 1985 1991 1996 2001 2006 2011 2016 Fishing and hunting

Wildlife watching3.

Level of reporting……...

Data collection agent….

NA

X

National

Private contrac-

tor

93 percent

X

National

U.S. Census Bureau

NA

X

National

U.S. Census Bureau

NA

X

National

U.S. Census Bureau

37 percent

X

State and National

Private contrac-

tor

90 percent

95 percent

State and

National

U.S. Census Bureau

92 percent

94 percent

State and

National

U.S. Census Bureau

95 percent

95 percent

State and

National

U.S. Census Bureau

80 percent

82 percent

State and

National

U.S. Census Bureau

88 percent

90 percent

State and

National

U.S. Census Bureau

77 percent

78 percent

State and

National

U.S. Census Bureau

67 percent

66 percent

State and

National

U.S. Census Bureau

67 percent

64 percent

National

U.S. Census Bureau

NA Not available. X Not applicable; wildlife watching (nonconsumptive) 1Spent $5.00 or more or participated 3 days or more during the year. 2Spent $7.50 or more or participated 3 days or more during the year. 3Termed “nonconsumptive” in 1980, 1985, and 1991 Surveys.

interviews were not conducted prior to 1980. Prescreening interview was introduced in 2016.

MAJOR CHANGES IN THE SURVEY

1955 to 1970 Surveys

The 1955 to 1970 Surveys included only substantial participants. Substantial participants were

defined as people who participated at least three days and/or spent at least $5.00 (the 1955–

1965 Surveys) or $7.50 (the 1970 Survey) during the surveyed year. Under most circumstances,

the Surveys may be compared for totals, but the effects of differences should be considered

when comparing the details of the Surveys.

The 1960, 1965, and 1970 Surveys differed from the 1955 National Survey in classification of

expenditures as outlined below.

1. Alaska and Hawaii were not included in the 1955 Survey.

2. Expenditure categories were more detailed in 1970 than in earlier Surveys.

3. The 1960 to 1970 classification of some expenditures differs from the 1955 Survey in the

following respects:

a. “Boats and boat motors” shown under “auxiliary equipment” were included in

“equipment, other” in 1955.

b. “Entrance and other privilege fees” asked separately were included in “trip

expenditures, other” in 1955.

c. “Snacks and refreshments” not included with “food” expenditures in the 1960 to

1970 reports were under “trip expenditures, other” in 1955.

d. Starting in 1960, expenditures on equipment, magazines, club dues, licenses, and

similar items were classified by the one sport activity for which expenditures

were chiefly made. In 1955, these expenditures were evenly divided among all

the activities in which the sportsperson took part.

e. Compared with 1955, the 1960 to 1970 Surveys reported fewer expenditures

within the “other” category because selected items were transferred to more

appropriate categories.

f. Expenditures on alcoholic beverages were reported separately in the 1970

Survey.


P a g e | 4

4. The number of waterfowl hunters in the 1970 Survey is not comparable with those

reported in the 1960 and 1965 Surveys. In 1960 and 1965, respondent sportspersons

were not included in the waterfowl hunter total if they reported that they went

waterfowl hunting but did not take the trip chiefly to hunt waterfowl. In 1970, all

respondents who reported that they had hunted waterfowl during 1970, regardless of

trip purpose, were included in the total. The number of hunters who did not take trips

chiefly to hunt waterfowl in 1970 was 1,054,000.

1975 Survey

In contrast to previous Surveys which covered substantial participants 12 years old and older,

the 1975 Survey based all the estimates on responses from individuals 9 years of age and older

and did not select respondents based upon substantial participation as defined above. As a

result, individuals who participated fewer than three days or spent less than $7.50 on hunting

or fishing were included in the estimates of participants, days of activity, and expenditures.

Categories of hunting and fishing expenditures differed from the previous four Surveys in that

only major categories were reported. For example, hunting equipment expenditures were not

further delineated by subcategory. Similarly, no detail was provided within the category of

fishing equipment expenditures. Expenses for items such as daily entrance fees, magazines,

club dues, and dogs were categorized as “other” in the 1975 report.

In addition to the above differences, the 1975 Survey gathered data on species sought for the

favorite hunting and fishing activity. This data replaced the “chiefly” category where hunting or

fishing was the primary purpose of the trip or day of activity. Data omitted in the 1975 Survey

that were included in previous Surveys include the respondents’ population density of

residence, occupation, and level of education.

1980 to 1985 Surveys

The 1980 and 1985 Surveys were similar. Each measured participants, rather than substantial

participants. Questions were incorporated into the 1980 and 1985 Survey questionnaires to

facilitate the construction of categories of data for comparisons with earlier Surveys. The use of

“chiefly” to delimit primary purpose appeared in the 1970 and prior Surveys, and its use was

continued in the 1980 and 1985 Surveys. The expenditure categories in 1980 and 1985 are

similar to the 1970 categories with the addition of fish finders, motor homes, and camper

trucks as separate categories. The definition of fishing included the use of nets or seines and

spearfishing. An extensive wildlife-watching section was added in 1980, necessitating a

separate detailed phase subsample.

As in the 1970 and 1975 Surveys, the 1980 and 1985 Surveys used a two-phase process to

gather information from households and individuals. In the first phase, household respondents


P a g e | 5

were asked to identify each participant six years of age and older who resided in their

household. In comparison, the 1975 and 1970 Surveys screened households for participants

who were nine years of age and older. In the second phase, the detailed interview phase,

interviews were conducted in person for the 1985, 1980, and 1970 Surveys and were conducted

by mail for the 1975 Survey. Participants were included in the detailed phase of the Survey if

they were at least 12 years old in 1970, 9 years old in 1975, and 16 years old in 1980 and 1985.

As a result, the population of hunters and anglers was more narrowly defined in 1980 and 1985.

However, estimates of sportspersons 6 years old and older, 9 years old and older, and 12 years

old and older, derived from the screening phase, are available for comparison with past

Surveys.

Overview of 1991 to 2016 Surveys Significant Methodological Differences

The most significant design differences in the six Surveys are as follows:

1. The 1991 Survey data was collected by interviewers filling out paper questionnaires. The data entries were keyed in a separate operation after the interview. The 1996, 2001, 2006, 2011, and 2016 Survey data were collected by the use of computer-assisted interviews. The questionnaires were programmed into computers, and the interviewer keyed in the responses at the time of the interview.

2. The 1991 Survey screening phase was conducted in January and February of 1991, when a household member of the sample households was interviewed on behalf of the entire household. The screening interviews for the 1996, 2001, and 2006 Surveys were conducted April through June of their survey years in conjunction with the first wave of the detailed interviews. The 2011 Survey also conducted screening interviews and the first detailed interviews April through June of 2011, but furthermore had an additional screening and detailed effort from February 2012 to the end of May 2012. The April–June 2011 screening effort had a high noncontact rate because of poor results using sample telephone numbers obtained from a private firm. The Census Bureau went back to the non-contacted component of the original sample in February–May 2012 and interviewed a subsample, requiring annual recall for those respondents. The Wave 3 screen sample was 12,484 of the total 48,600 household screen sample. A modification of the 2011 sampling scheme was to oversample counties that had relatively high proportions of hunting license purchases. The screening interviews for the 2016 Survey were conducted April 1, 2016 through May 15, 2016 in conjunction with the first wave of the detailed interviews.

The screening interviews for all six Surveys consisted primarily of demographic questions and wildlife-related recreation questions concerning activity in the previous year (1990, 1995, etc.) and intentions for recreating in the survey year. In the 1991 Survey, an attempt was made to contact every sample person in all three detailed interview waves. In 1996, 2001, 2006, 2011, and 2016 respondents who were interviewed in the first detailed interview wave were not contacted again until the third wave (unless they


P a g e | 6

were part of the other subsample, i.e., a respondent in both the sportsperson and wildlife-watching subsamples could be in the first and third wave of sportsperson interviewing and the second and third wave of wildlife-watching interviewing). Also, all interviews in the second wave were conducted only by telephone. In-person interviews were only conducted in the first and third waves. The 2011 and 2016 Wave 3 screen phases were composed of both telephone and in-person interviews.

1991 Survey

A significant change was made in 1991 in the recall period used in the detailed phase of the

FHWAR Surveys. The recall period in 1991 was shortened from the 12 months used in previous

Surveys to 4 months in order to improve the accuracy of the data collected. As a result of that

change, the Surveys conducted since 1991 cannot be compared with those conducted earlier.

The 1955 to 1985 Surveys required respondents to recall their recreation activities for the

survey year at the beginning of the following year. The 1991 to 2016 Surveys went to the

respondents two or three times during the survey year to get their activity information. The

change in the recall period was due to a study1 of the effect of the respondent recall length on

Survey estimates. The study found significant differences in FHWAR survey results using annual

recall periods versus shorter recall periods. Longer recall periods lead to higher estimates. Even

when everything else was held constant, such as questionnaire content and sample design,

increasing the respondent’s recall period resulted in significantly higher estimates for the same

phenomenon.

The recall study also found that the extent of recall bias varied for different types of fishing and

hunting participation and expenditures. For example, annual recall respondents gave an

estimate of average annual days of saltwater fishing that was 46 percent higher than the

trimester recall estimate, while the annual recall estimate of average annual saltwater fishing

trips was 30 percent higher than the trimester recall estimate. This means there is no single

correction factor for all survey estimates when calculating trends from surveys using different

recall periods. Reliable trends analysis needs to use data compiled from surveys in which the

important elements, such as the sample design and recall period, are not significantly different.

1996 Survey

1. The 1991 Survey collected information on all wildlife-related recreation purchases made by participants without reference to where the purchase was made. The 1996 Survey asked in which state the purchase was made.

2. In 1991, respondents were asked what kind of fishing they did, i.e., Great Lakes, other freshwater, or saltwater, and then were asked in what states they fished. In 1996, respondents were asked in which states they fished and then were asked what kind of

1Investigation of Possible Recall/Reference Period Bias in National Surveys of Fishing, Hunting and Wildlife-Associated Recreation, December 1989, Westat, Inc.


P a g e | 7

fishing they did. This method had the advantage of not asking about, for example, saltwater fishing when they only fished in a noncoastal state.

3. In 1991, respondents were asked how many days they “actually” hunted or fished for a particular type of game or fish and then how many days they “chiefly” hunted or fished for the same type of game or fish rather than another type of game or fish. To get total days of hunting or fishing for a particular type of game or fish, the “actually” day response was used, while to get the sum of all days of hunting or fishing, the “chiefly” days were summed. In 1996, respondents were asked their total days of hunting or fishing in the country and each state, then how many days they hunted or fished for a particular type of game or fish.

4. Trip-related and equipment expenditure categories were not the same for all Surveys. “Guide fee” and “Pack trip or package fee” were two separate trip-related expenditure items in 1991, while they were combined into one category in the 1996 Survey. “Boating costs” was added to the 1996 hunting and wildlife-watching trip-related expenditure sections. “Heating and cooking fuel” was added to all of the trip-related expenditure sections. “Spearfishing equipment” was moved from a separate category to the “other” list. “Rods” and “Reels” were two separate categories in 1991 but were combined in 1996. “Lines, hooks, sinkers, etc.” was one category in 1991 but split into “Lines” and “Hooks, sinkers, etc.” in 1996. “Food used to feed other wildlife” was added to the wildlife-watching equipment section, “Boats” and “Cabins” were added to the wildlife-watching special equipment section, and “Land leasing and ownership” was added to the wildlife-watching expenditures section.

5. Questions asking sportspersons if they participated as much as they wanted were added in 1996. If the sportspersons said no, they were asked why not.

6. The 1991 Survey included questions about participation in organized fishing competitions; anglers using bows and arrows, nets or seines, or spearfishing; hunters using pistols or handguns and target shooting in preparation for hunting. These questions were not asked in 1996.

7. The 1996 Survey included questions about catch and release fishing and persons with disabilities participating in wildlife-related recreation. These questions were not part of the 1991 Survey.

8. The 1991 Survey included questions about average distance traveled to recreation sites. These questions were not included in the 1996 Survey.

9. The 1996 Survey included questions about the last trip the respondent took. Included were questions about the type of trip, where the activity took place, and the distance and direction to the site visited. These questions were not asked in 1991.

10. The 1991 Survey collected data on hunting, fishing, and wildlife-watching by U.S. residents in Canada. The 1996 Survey collected data on fishing and wildlife-watching by U.S. residents in Canada.

2001 Survey

1. The 1991 and 1996 single race category “Asian or Pacific Islander” was changed to two

categories “Asian” and “Native Hawaiian or Other Pacific Islander.” In 1991 and 1996,


P a g e | 8

the respondent was required to pick only one category, while in 2001 the respondent

could pick any combination of categories. The next question stipulated that the

respondent could only be identified with one category and then asked what that

category was.

2. The 1991 and 1996 land leasing and ownership sections asked the respondent to

combine the two types of land use into one and give total acreage and expenditures. In

2001, the two types of land use were explored separately.

3. The 1991 and 1996 wildlife-watching sections included questions on birdwatching for

around-the-home participants only. The 2001 Survey added a question on birdwatching

for away-from-home participants. Also, questions on the use of birding life lists and how

many species the respondent can identify were added.

4. “Recreational vehicles” was added to the sportspersons and wildlife-watching special

equipment section. “House trailer” was added to the sportspersons special equipment

section.

5. Total personal income was asked in the detailed phase of the 1996Survey. This was

changed to total household income in the 2001 Survey.

6. A question was added to the trip-related expenditures section to ascertain how much of

the total was spent in the respondent’s state of residence when the respondent

participated in hunting, fishing, or wildlife-watching out-of-state.

7. Boating questions were added to the fishing section. The respondent was asked about

the extent of boat usage for the three types of fishing.

8. The 1996 Survey included questions about the months around-the-home wildlife-

watching participants fed birds. These questions were not repeated in the 2001 Survey.

9. The contingent valuation sections of the three types of wildlife-related recreation were

altered, using an open-ended question format instead of 1996’s dichotomous choice

format.

2006 Survey

1. A series of boating questions was added. The new questions dealt with anglers using

motorboats and/ or non-motorboats, length of boat used most often, distance to boat

launch used most often, needed improvements to facilities at the launch, whether or

not the respondent completed a boating safety course, who the boater fished with most

often, and the source and type of information the boater used for his or her fishing.

2. Questions regarding catch and release fishing were added. They were whether or not

the respondent caught and released fish and, if so, the percent of fish released.

3. The proportion of hunting done with a rifle or shotgun, as contrasted with muzzleloader

or archery equipment, was asked.

4. In the contingent valuation section, where the value of wildlife-related recreation was

determined, two quality-variable questions were added: the average length of certain

fish caught and whether a deer, elk, or moose was killed. Plus the economic evaluation


P a g e | 9

bid questions were rephrased, from “What is the most your [species] hunting in [State

name] could have cost you per trip last year before you would NOT have gone [species]

hunting at all in 2001, not even one trip, because it would have been too expensive?”,

for the hunters, for example, to “What is the cost that would have prevented you from

taking even one such trip in 2006? In other words, if the trip cost was below this

amount, you would have gone [species] hunting in [State name], but if the trip cost was

above this amount, you would not have gone.”

5. Questions concerning hunting, fishing, or wildlife-watching in other countries were

taken out of the Survey.

6. Questions about the reasons for not going hunting or fishing, or not going as much as

expected, were deleted.

7. Disability of participants questions were taken out.

8. Determination of the types of sites for wildlife-watching was discontinued.

9. The birding questions regarding the use of birding life lists and the ability to identify

birds based on their sight or sounds were deleted.

10. Public transportation costs were divided into two sections, “public transportation by

airplane” and “other public transportation, including trains, buses, and car rentals, etc.”.

2011 Survey

1. The series of boating questions added in 2006 was deleted.

2. Questions about target shooting and the usage of a shooting range in preparation for

hunting were added. The types of weapon used at the shooting range were quantified.

3. Questions about plantings expenditures for the purpose of hunting were added.

4. “Feral pig” was re-categorized from big game to other animals for all states except

Hawaii.

5. “Ptarmigan” was included as its own small game category, instead of lumped in “other.”

6. In previous Surveys, “Moose” was included as its own category only for Alaska. For

2011, “Moose” was included as its own big game category, instead of lumped in “other,”

for all 50 states.

7. In previous Surveys, “Wolf” was included as its own category only for Alaska. For 2011,

“Wolf” was included as its own other animal category, instead of lumped in “other,” for

all 50 states.

8. The household income categories were modified. The top categories were changed

from “$100,000 or more” to “$100,000 to $149,999” and “$150,000 or more.”

9. The “Steelhead” category was deleted from the saltwater fish species section, with the

idea that it would be included in “other.”

10. The 2006 around-the-home wildlife-watching category that quantified visitors of “public

parks or areas” was rewritten to wildlife-watching at “parks or natural areas.” This

change was to make clear that respondents should include recreating at

quasigovernmental and private areas.


P a g e | 10

11. The 2006 wildlife-watching equipment category “Film and developing” was rewritten to

“Film and photo processing.”

2016 Survey

1. Recreational archery and target shooting with firearms questions were added to the

screening instrument. These questions were not asked only of hunters; they were general

population questions.

2. The around-the-home wildlife-watching questions in the screening instrument were

changed from asking about four types of wildlife watching (observing, photographing,

feeding, and maintaining natural areas or plantings for the benefit of wildlife) to asking one

question (wildlife watching around the home).

3. The contingent valuation questions were deleted. These were the valuation questions for

moose, elk, and deer hunting, walleye, trout, and black bass fishing, and away-from-home

wildlife watching.

4. The questions in the special equipment section asking if the respondent would have bought

the item if they had not gone hunting, fishing, or wildlife watching were deleted.

5. The screening instrument was redesigned to ask the 2016 participation of household

members 16 years and older at the beginning of the interview. If the household member

participated in 2016, the rest of the activity section in the screener, which covered

participation in 2015, was skipped. The household member was selected for the detailed

interview in the case of fishing and hunting. For wildlife watching, the household member

was eligible for selection for the detailed interview.

Design of Survey Sample 2016 FHWAR Design and Methodology Technical Report

P a g e | 11

Chapter 3. DESIGN OF THE NATIONAL SURVEY OF FISHING, HUNTING, AND WILDLIFE ASSOCIATED RECREATION SURVEY SAMPLE

INTRODUCTION

The 2016 FHWAR was designed to provide national-level estimates of the number of

participants in recreational hunting and fishing and in wildlife-watching activities (i.e., wildlife

observation). Information was collected on the number of participants, where and how often

they participated, the type of wildlife encountered, and the amounts of money spent on

wildlife-related recreation.

The requirements set forth by AFWA for the 2016 FHWAR dictated that the U.S. Census Bureau

collect estimates of fishing, hunting and wildlife-associated activities at the national level as

well as for four states (Maine, Minnesota, Oklahoma, and Virginia). In order to achieve

statistically sound data outcomes for these geographies within budget, the Census Bureau

redesigned the FHWAR sample to be more efficient.

Similar to FHWAR surveys from the past, the 2016 FHWAR required a multistage probability

sample that represented all 50 states and Washington, D.C. (a national sample design). The first

stage of the survey design involved the formation, stratification, and selection of primary

sampling units (PSUs). The second stage involved sampling housing units (HUs) from the Census

Bureau’s Master Address File (MAF). For the 2016 sample design, the Census Bureau defined

the PSUs using information related to hunting rather than using the Current Population Survey’s

(CPS) PSUs, which had been used in the past. The 2016 FHWAR PSUs targeted high hunter

participation areas with the goals of minimizing the variance and minimizing field

representatives’ (FR) travel costs.

The past FHWAR surveys were designed to produce state-level estimates for each of the 50

states and the national-level estimates. In 2016, the FHWAR survey was designed to produce

national-level estimates that are representative of the entire United States and state-level

estimates for only Maine, Minnesota, Oklahoma, and Virginia. Thus, the Census Bureau defined

the PSUs within these four states and the nine census divisions. The PSUs defined within the

four states did not cross state boundaries; however, PSUs within a division and not in the four

designated states could cross state boundaries, but did not cross division boundaries.

After sampling, the survey was conducted in three phases: a pre-screener sample, an initial

screening of households to identify likely sportspersons and wildlife-watching participants, and

a series of follow-up interviews of selected household members to collect detailed data about

their wildlife-related recreation during 2016.


P a g e | 12

The target population for the 2016 FHWAR was the household population, which was similar to

the 2011 FHWAR. The sampling frame consisted of all valid HUs in the July 2015 MAF.

THE MASTER ADDRESS FILE (MAF)

The Census Bureau’s demographic surveys, including FHWAR, select their samples from two

dynamic sampling frames, one for HUs and one for group quarters (GQs)2, which are based on

the MAF. The MAF is a national inventory of addresses that is continually updated by the

Census Bureau to support its decennial programs and demographic surveys. The MAF is

maintained by the Census Bureau’s Geography Division (GEO).

The MAF replaced a variety of address sources used in the past to construct sampling frames

for the demographic surveys. For the sample design based upon the 2000 and earlier censuses,

the demographic survey samples were selected from a coordinated set of four sampling frames:

the unit frame, the area frame, the GQ frame, and the permit new construction frame. The

address sources for these frames included the official address list from the most recent census,

block listings, and addresses from building permits.

Creating and Updating The Demographic Sampling Frames

The current HU Frame was created for the first time in 2013 and is updated every six months

with the latest MAF data. GEO delivers MAF extracts twice each year, in January and July. A

MAF extract is a “snapshot” of the MAF for a given county that reflects six months of Delivery

Sequence File (DSF) and other updates.

The MAF filtering is a critical feature of the frame creation process; its outcomes can have an

important effect on frame coverage. The MAF extracts contain all records from the MAF for a

given county, including many that should not be eligible for the FHWAR HU Frame. The filtering

rules designate each MAF record as either “valid” (passed the filter and eligible for the HU

Frame) or “invalid” (failed the filter, ineligible for the Frame).

The FHWAR HU Frame takes the form of separate HU universes by county, just as the MAF

extracts are separate by county. The HU Frame files are called the Unit Frame Universe Files

(UFUFs). The original UFUFs for demographic surveys were created in 2013 and consisted of all

the valid and invalid HUs from the MAF at that time. For FHWAR, only valid HUs from the July

2015 MAF were included in the UFUFs since the sample is a one-time sampling operation.

Starting with those initial 2013 universe files, the UFUFs are updated every six months with

MAF data in two ways:

1. Each existing UFUF record is updated with the most recent MAF data (addresses,

block codes, etc.) and its latest filtering status.

2People living in GQs were not in-scope for FHWAR.


P a g e | 13

2. New growth records are added to the UFUF.

The UFUFs are also updated with sort information from the American Community Survey (ACS)

and decennial block-level data as part of the annual sampling process, which takes place once

each year as part of the January MAF processing cycle. Each survey participating in annual

sampling can sort the frame units in its own way.

ADMINISTRATIVE DATA

To target high hunter participation areas, auxiliary data related to hunting were needed. With

assistance from FWS, the Census Bureau received resident hunter license-holder counts for

every county, parish, borough, and municipal area in 49 of 50 states from the states’ wildlife

agencies.3,4 FWS requested that the individual state send the most up-to-date data available.

West Virginia was unable to produce such hunter counts; however, for the 2011 FHWAR, West

Virginia did provide the Census Bureau the hunter counts.5 The Census Bureau made the

decision to use the 2011 FHWAR hunter counts in West Virginia for the 2016 FHWAR sample

design under the assumption that the numbers may not have changed much from 2011 and the

counts were the best counts available for the state.

SAMPLE SIZE

2013 FHWAR Pre-Screener Test

The Census Bureau used a pre-screener, self-response questionnaire for the 2016 FHWAR, to

determine whether the household members were eligible to participate in the follow-up detail

questionnaire. Results from the 2013 FHWAR Pre-Screener Test were used to predict the

response rates for the 2016 FHWAR pre-screener. Table 3.1 shows the pre-screener test

response and nonparticipation rates used for the sample size calculations.

Table 3.1 2013 Pre-Screener Test Results

Response Status Results Rates

Non Respondent No Return 60.97 Respondent Nonparticipant 13.07 Respondent Participant with phone numbers 17.60 Undeliverable Undeliverable As Addressed (UAA's) 8.36

3State wildlife agencies provided resident hunter license-holder counts for the calendar year or the 12 months prior to purchase of the license for 2012, 2013 or 2014. 4Hunter counts were not requested from Washington, D.C. since hunting is not allowed in Washington, D.C. The sample in Washington, D.C. was included with certainty to have representation of the whole country. 5For the 2011 FHWAR, Census requested and received the same data from all 50 states, which was used to over/under sample HUs in the CPS defined sample areas to target hunters.


P a g e | 14

The nonparticipant and the phone number groups both returned their pre-screener test

questionnaires, which amounted to approximately a 31% response rate.6 The pre-screener test

results, the coefficient of variation (CV) requirements, and the cost estimates were used in

combination when producing the total pre-screener sample size.

National Pre-Screener Sample Size

The 2016 FHWAR strived for a minimum of 8,000 households to interview for the screener

questionnaire, which would result in an estimated CV of 8% for the national hunter estimate. In

2011, the national hunter CV was much lower at 3%. Due to 2016 FHWAR cost constraints, a 3%

CV level was determined to be unreachable.

The formula used to calculate the estimated hunter CV was:

𝐶𝑉ℎ𝑢𝑛𝑡𝑒𝑟 = √(1 − 𝑝) ∗𝐷𝐸𝐹𝐹

𝑛 ∗ 𝑝

where

𝑝 = 𝑝𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓ℎ𝑢𝑛𝑡𝑒𝑟𝑠 (0.06)

𝑛 = 𝑠𝑐𝑟𝑒𝑒𝑛𝑒𝑟 𝑠𝑎𝑚𝑝𝑙𝑒 𝑠𝑖𝑧𝑒

the proportion of hunters was 0.06 (from the 2011 FHWAR),

and the design effect (DEFF) was calculated as:

𝐷𝐸𝐹𝐹 = 𝐷𝐸𝐹𝐹𝑠𝑢𝑏𝑠𝑎𝑚𝑝𝑙𝑒 𝐷𝐸𝐹𝐹 ∗ 𝐷𝐸𝐹𝐹𝑛𝑎𝑡𝑖𝑜𝑛𝑎𝑙 𝐷𝐸𝐹𝐹

where

𝐷𝐸𝐹𝐹𝑠𝑢𝑏𝑠𝑎𝑚𝑝𝑙𝑒 𝐷𝐸𝐹𝐹 =𝑛 ∗ ((𝑛𝐶𝐴𝑃𝐼 ∗ (𝑆𝐹𝐶𝐴𝑃𝐼 ∗ 𝑆𝐼)2) + (𝑛𝐶𝐴𝑇𝐼 ∗ (𝑆𝐹𝐶𝐴𝑇𝐼 ∗ 𝑆𝐼)2))

((𝑛𝐶𝐴𝑃𝐼 ∗ 𝑆𝐹𝐶𝐴𝑃𝐼 ∗ 𝑆𝐼) + (𝑛𝐶𝐴𝑇𝐼 ∗ 𝑆𝐹𝐶𝐴𝑇𝐼 ∗ 𝑆𝐼))2

𝑆𝐹 = 𝑠𝑢𝑏𝑠𝑎𝑚𝑝𝑙𝑖𝑛𝑔 𝑓𝑎𝑐𝑡𝑜𝑟

𝑆𝐼 = 𝑆𝑎𝑚𝑝𝑙𝑖𝑛𝑔 𝐼𝑛𝑡𝑒𝑟𝑣𝑎𝑙 =𝐽𝑢𝑙𝑦 2013 𝑀𝐴𝐹 𝐻𝑜𝑢𝑠𝑖𝑛𝑔 𝑈𝑛𝑖𝑡 𝐶𝑜𝑢𝑛𝑡

𝑛

6Respondents were given the option to call the Census Contact Center with questions or to complete the interview. The U.S. Government was shut down for two and a half weeks in the middle of the pre-screener test. The effects of the shutdown on the test are unknown, but should be taken into consideration when reviewing the results.


P a g e | 15

To achieve 4,000 CATI cases with valid phone numbers, the pre-screener sample size was

determined to be 22,725.

𝑃𝑆𝑆 ∗ 0.1760 = 4,000

With a sample size of 22,725, this leads to an estimated 13,855 nonrespondent households. A

subsample of 1 out of every 3.4635 households would provide 4,000 CAPI cases.

22,725 ∗ 0.6097 = 13,855 13,855

𝑆𝐹𝐶𝐴𝑃𝐼= 4,000

Recall from Table 3.1, 0.1760 and 0.6097 were the proportions of the respondents who

provided phone numbers and of nonrespondents for the 2013 Pre-Screener Test, respectively.

The total estimated screener sample size would then be 8,000 households: 4,000 CAPI and

4,000 CATI.

Four States’ Sample Size Calculations

The Census Bureau was required to produce state-level estimates for four states: Maine,

Minnesota, Oklahoma, and Virginia. The requirement for these four states was that the hunter

CV needed to be no larger than 15%. Using past data and generalized variance parameters, the

Census Bureau calculated the sample sizes needed for the four states.

For each of the four states, the following steps were used to determine state samples sizes:

1) The 2011 state CV was calculated using 2011 generalized variance function a and b

parameters:

𝑆𝐸𝑠𝑡𝑎𝑡𝑒 = √𝑎 ∗ ℎ𝑢𝑛𝑡𝑒𝑟𝑠𝑠𝑡𝑎𝑡𝑒2 + 𝑏 ∗ ℎ𝑢𝑛𝑡𝑒𝑟𝑠𝑠𝑡𝑎𝑡𝑒

𝐶𝑉𝑠𝑡𝑎𝑡𝑒 =𝑆𝐸𝑠𝑡𝑎𝑡𝑒

ℎ𝑢𝑛𝑡𝑒𝑟𝑠𝑠𝑡𝑎𝑡𝑒

where SEstate is the standard error (SE) for the state estimate of the number of hunters,

and Huntersstate is the state estimate of the number of hunters.

2) The estimated number of 2016 cases to be completed using the 2011 completed

screeners, the 2011 CV, and the required 2016 CV was calculated:


P a g e | 16

2016 𝐶𝑎𝑠𝑒𝑠 𝑡𝑜 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒 = (2011 𝐶𝑉𝑠𝑡𝑎𝑡𝑒

𝑇𝑎𝑟𝑔𝑒𝑡 2016 𝐶𝑉𝑠𝑡𝑎𝑡𝑒)2 ∗ 2011 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑 𝑆𝑐𝑟𝑒𝑒𝑛𝑒𝑟𝑠

3) The number of 2016 cases to assign were calculated as:

2016 𝐴𝑠𝑠𝑖𝑔𝑛𝑒𝑑 𝐶𝑎𝑠𝑒𝑠 =2016 𝐶𝑎𝑠𝑒𝑠 𝑡𝑜 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒

2011 𝑅𝑎𝑡𝑖𝑜 𝑜𝑓 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒 𝑡𝑜 𝐴𝑠𝑠𝑖𝑔𝑛𝑒𝑑

4) The expected 2016 CV was calculated by starting with a ratio of estimated completed

screener interviews in 2016 to the complete screeners in 2011 and calculating the

estimated 2016 a and b parameters as:

𝑎 =2011 𝑎

2016 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑 𝐶𝑎𝑠𝑒𝑠2011 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒 𝐶𝑎𝑠𝑒𝑠⁄

𝑏 =2011 𝑏

2016 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑 𝐶𝑎𝑠𝑒𝑠2011 𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒 𝐶𝑎𝑠𝑒𝑠⁄

With these “new” a and b parameters the 2016 standard error (SE) for hunters was

calculated with the assumption that the total number of hunters per state (huntersstate)

would be the same as 2011. With the 2016 SE, the 2016 CV could also be calculated:

2016 𝑆𝐸𝑠𝑡𝑎𝑡𝑒 = √2016 𝑎 ∗ ℎ𝑢𝑛𝑡𝑒𝑟𝑠𝑠𝑡𝑎𝑡𝑒2 + 2016 𝑏 ∗ ℎ𝑢𝑛𝑡𝑒𝑟𝑠𝑠𝑡𝑎𝑡𝑒

2016 𝐶𝑉𝑠𝑡𝑎𝑡𝑒 =2016 𝑆𝐸𝑠𝑡𝑎𝑡𝑒

ℎ𝑢𝑛𝑡𝑒𝑟𝑠𝑠𝑡𝑎𝑡𝑒

Census Divisions’ Sample Size Calculations

The formulas used to calculate the Census divisions’ sample sizes are the same formulas used to

calculate the four states’ sample sizes. The a and b parameters used are not published

parameters; these were calculated specifically for this sample size exercise.


P a g e | 17

Table 3.2 Sample Sizes and Expected CVs for the Sample Areas

Sample Area Code Sample Area States

Sample Size

Hunter CV7

01 New England Connecticut, Massachusetts, New Hampshire, Rhode Island, Vermont

4,106 0.10

02 Middle Atlantic New Jersey, New York, Pennsylvania 2,128 0.10

03 East North Central

Illinois, Indiana, Michigan, Ohio, Wisconsin 1,315 0.10

04 West North Central

Iowa, Kansas, Missouri, Nebraska, North Dakota, South Dakota

1,123 0.10

05 South Atlantic Delaware, Florida, Georgia, Maryland, North Carolina, South Carolina, Washington, D.C., West Virginia

2,104 0.10

06 East South Central

Alabama, Kentucky, Mississippi, Tennessee 675 0.10

07 West South Central

Arkansas, Louisiana, Texas 868 0.10

08 Mountain Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, Wyoming

1,640 0.10

09 Pacific Alaska, California, Hawaii, Oregon, Washington

3,015 0.10

23 Maine Maine 796 0.15

27 Minnesota Minnesota 352 0.15

40 Oklahoma Oklahoma 1,210 0.15

51 Virginia Virginia 3,393 0.15

Total 22,725

FORMATION AND SAMPLE SELECTION OF PSUs

Sample Areas

For the 2016 FHWAR, the Census Bureau was required to produce the national fishing, hunting,

and wildlife-watching estimates along with individual state estimates for Maine, Minnesota,

Oklahoma, and Virginia. The prerequisite HUs sample sizes for the four states were required to

result in estimated CVs no larger than 15% for the number of hunters. The other forty-six states

and Washington, D.C. received the remaining sample cases to arrive at a national sample of

22,725 HUs.8

7The estimated weighted CV for all sample areas was 0.08. 8The four states’ sample sizes are included in the national estimates.


P a g e | 18

Figure 3.1 shows the four Census regions and nine Census divisions. The Census Bureau divided

the balance of the United States into the nine census divisions, resulting in thirteen sample

areas as represented in Table 3.2.

Figure 3.1 Census Regions and Divisions

Sample Area Preparation

Within each sample area, the overall sampling interval (SI) was calculated as the ratio of the

sample area’s total 2015 Valid Housing Units (VHU) and sample area’s sample size.

𝑆𝑎𝑚𝑝𝑙𝑖𝑛𝑔 𝐼𝑛𝑡𝑒𝑟𝑣𝑎𝑙 (𝑆𝐼) =𝑇𝑜𝑡𝑎𝑙 2015 𝑉𝐻𝑈

𝑆𝑎𝑚𝑝𝑙𝑒 𝑆𝑖𝑧𝑒

Percent license and expected sample size were calculated at the county level as follows:

𝑃𝑒𝑟𝑐𝑒𝑛𝑡 𝐿𝑖𝑐𝑒𝑛𝑠𝑒 =𝐻𝑢𝑛𝑡𝑒𝑟 𝐶𝑜𝑢𝑛𝑡

2013 𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝐶𝑜𝑢𝑛𝑡∗ 100

𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑆𝑎𝑚𝑝𝑙𝑒 𝑆𝑖𝑧𝑒 =𝑇𝑜𝑡𝑎𝑙 2015 𝑉𝐻𝑈

𝑆𝐼


P a g e | 19

Percent license shows the percentage of the population per county that obtained a hunting

license. The expected sample size shows the number of potential VHUs sampled for the pre-

screener in each county, as the sample is proportional to sample size. These values were used

in the formation of the PSUs.

Defining the PSUs

Defining of the PSUs occurred in each of the thirteen sample areas. Both self-representing (SR)

and non-self-representing (NSR) PSUs were defined for all areas. SR PSUs consisted of individual

counties, while NSR PSUs contained one or more counties.9

Self-Representing (SR) PSUs

A combined score was formulated to assist with the determination of counties to be SR PSUs.

The combined score was comprised of the county’s number of hunters, percentage of hunters,

and 2015 VHU count. The counties within a sample area were ranked 1, 2, …,n, where n =

number of counties, for each of the three variables: number of hunters, percentage of hunters,

and 2015 VHU counts. The combined score was calculated as follows:

𝐶𝑜𝑚𝑏𝑖𝑛𝑒𝑑 𝑆𝑐𝑜𝑟𝑒 = 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐻𝑢𝑛𝑡𝑒𝑟𝑠 𝑅𝑎𝑛𝑘 ∗ 0.30 +

𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝐿𝑖𝑐𝑒𝑛𝑠𝑒 𝑅𝑎𝑛𝑘 ∗ 0.25 +

𝑉𝐻𝑈 𝑅𝑎𝑛𝑘 ∗ 0.45

The proportions of 0.30, 0.25, and 0.45 were used for number of hunters rank, percentage

license rank, and VHU rank, respectively, for SR PSU selection were chosen for several

reasons.10 The percentage of licenses was important to the sample design but, in many cases,

counties with higher percentages of licenses also had lower population counts, particularly in

rural counties. The Census Bureau would have liked to include all high-percentage license

counties for sampling purposes, but having a county with an expected sample size of less than

one HU would be costly and inefficient. This is why the VHU rank proportion was higher, at

0.45. The Census Bureau needed to find the hunters, but also needed to accurately represent

the sample area. The number of hunters was also an important measure of hunting

participation and was included in the combined score accordingly.

The combined score was then sorted lowest to highest. The lower the combined score, the

more likely the PSU would be a SR PSU. Ward’s hierarchical clustering method was performed

9Bedford County and Bedford City, Virginia were an exception and formed an SR PSU because the Census Bureau did not receive hunter counts from Bedford City. An assumption was made that the counts were combined within Bedford County. 10The West North Central division used the following formula for combined score, due to small expected sample sizes: Combined Score = number of hunter rank * 0.2 + percentage of hunter rank * 0.15 + VHU rank * 0.65


P a g e | 20

on the combined score using the interactive software JMP.11 The number of PSUs to select as

SR is not an exact science. The number of SR counties in the 2010 CPS design was looked at as a

guideline, as well as considering the total sample size per sample area and the hunter activity

distribution.

Figure 3.2 contains a colored dendrogram and distance graph to represent the results of the

clustering analysis for a made-up example. This output, from the Ward’s hierarchical clustering,

helps guide the selection of a meaningful number of clusters. It is an acceptable practice to pick

a point after the line has been flat and right before the graph’s line shoots up (from left to right)

on the dendrogram as the criteria to identify the SR PSUs. For this example, the first two

clusters, made up of 10 counties, were identified as SR PSUs. The SR PSUs were formed from

these ten individual counties and selected with certainty into sample. When forming the non-

self-representing PSU, these counties were removed.

In Figure 3.2, the counties within the clusters were formed to be as homogeneous as possible

based on the Ward’s hierarchical clustering. The lines, moving from left to right, indicate the

homogeneity of the paired counties. In this example, counties 3 and 4 are very homogeneous;

whereas, county 27 is not as homogeneous as the others, and thus, is clustered by itself.

11JMP is an interactive software developed by the SAS institute.


P a g e | 21

Figure 3.2 Hierarchical Clustering Dendogram

Non-Self-Representing (NSR) PSUs and Formation of Strata

The NSR PSU creation method included many steps. Using JMP’s mapping software and a list of

criteria, strata and NSR PSUs for each sample area were formed.

The geographical makeup of the strata were vital to the FHWAR for a number of reasons. The

FHWAR asked specific questions on species of wildlife hunted, fished, and watched. If the

stratum contained counties that were far apart from each other or had different types of


P a g e | 22

environmental makeup, then valuable information could be lost. For example, suppose three

counties border saltwater and two other counties within the same stratum are mountainous. If

a mountainous county were selected, the potential saltwater anglers could be missed

altogether. Thus, in this example, the five counties should be put into two different strata:

three saltwater counties in one stratum and the two mountainous counties in a different

stratum.

Criteria and Priorities

The following five objectives, in priority order, were set to assist with the PSU selection and

stratification:

1st Objective - Minimize the first stage component of variance (within strata, between

PSUs); the variables of interest are number of hunters and percentage of

hunters

Between NSR strata hunter/percent score to be as heterogeneous as

possible.

Hunter/percent score calculated as:

(𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐻𝑢𝑛𝑡𝑒𝑟𝑠 𝑅𝑎𝑛𝑘 ∗ 0.5 + 𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝐿𝑖𝑐𝑒𝑛𝑠𝑒 𝑅𝑎𝑛𝑘 ∗ 0.50)12

If possible, NSR PSUs within strata to be as homogeneous as possible,

with respect to the hunter/percent score and geographic make up.

2nd Objective - At least 15 expected sample HUs per stratum to make FR workloads more

efficient.

3rd Objective - Similar size NSR strata, in terms of the number of VHUs, to minimize

between strata variance.

4th Objective - Counties of the same PSU need to be contiguous. NSR PSUs can be either

single-county or a group of contiguous counties. PSUs should be within

similar geographical areas.

5th Objective - Roughly equal PSU VHU sizes within a stratum to attempt to have equal

probabilities of selection of PSUs.

Another rule that was applied when forming NSR PSUs was no PSU could be over 3,000 square

miles (unless a single county is over 3,000 square miles). This rule was also used in the redesign

of demographic surveys to minimize FR travel costs.

12The hunter/percent score is a measurement to assist with the mapping and identification of hunter activity. 0.5 was used because the number and percentage of hunters is equally important for forming the NSR PSUs.


P a g e | 23

PSU Sample Selection

The first stage of the two-stage probability sample involves selecting the PSUs.13 All of the SR PSUs were selected with certainty. For the NSR PSUs, one PSU was selected per stratum. The 2016 FHWAR NSR PSUs were selected with probability proportional to the size of the PSUs, where the measure of size is the PSU housing unit count.

Sample Design Summary Tables

Table 3.3 gives the number of strata, PSUs, and counties for the overall PSU sample design and breaks each category down to SR and NSR. One SR PSU in Virginia contains two counties – Bedford County and the City of Bedford,14 thus the number of counties is one greater than the number of SR PSUs in the tables below. Table 3.3 also provides the number of PSUs and counties in the 2016 FHWAR sample. Notice 753 PSUs were selected, one from each stratum.

Table 3.3 Sample Universe and In-Sample Counts

Sample Universe In-Sample

Strata 753 N/A15 SR 540 N/A

NSR 213 N/A

PSU 2,013 753 SR 540 540

NSR 1,473 213

Counties 3,143 945 SR 541 541

NSR 2,602 404

13The second stage was sampling housing units within the selected PSUs. 14Independent cities are treated as counties. 15N/A Not Applicable. There is no “in-sample” for strata, as the sample was selected in PSU and counties.


P a g e | 24

Table 3.4 provides a breakdown of the sample and the number of SR/NSR PSUs in each sample area.

Table 3.4 Sample Universe, Sample Areas’ Strata, County, and PSU Counts with In-Sample SR/NSR Counts

Sample Area Strata County Primary

Count Sampling

SR Unit

NSR

01

02

03

04

05

06

07

08

09

–

–

–

–

–

–

–

–

–

New England

Middle Atlantic

East North Central

West North Central

South Atlantic

East South Central

West South Central

Mountain

Pacific

21

49

103

60

110

79

93

72

57

51

150

437

531

455

364

393

281

167

44

110

233

291

245

175

270

273

159

9

29

78

36

73

59

83

56

44

12

20

25

24

37

20

10

16

13

23

27

40

51

–

–

–

–

Maine

Minnesota

Oklahoma

Virginia

10

24

24

51

16

87

77

134

14

56

51

92

6

18

14

35

4

6

10

16

Total 753 3,143 2,013 540 213

PRE-SCREENER SAMPLE: SAMPLE SELECTION OF HOUSEHOLDS WITHIN PSUs

The second-stage of FHWAR sample design was selecting valid HUs within the selected PSUs from the MAF/UFUF as defined for FHWAR. The HUs within the PSU UFUFs were selected using a systematic sample procedure. Within each sample area, the HUs were ordered by PSU number, Federal Information Processing Standards (FIPS) state code, FIPS county code, geocode flag, current block state, current block county, current block tract, current block code, ZIP code, and Master Address File Identification (MAFID) number. A sample area sampling interval (SI) was calculated using the housing unit count within the sampling area divided by the total sample size for the sample area. The sample area was then multiplied by the probability of the PSU that was selected into sample to determine the within-PSU SI. This step ensured a self-weighting design in which each sampled HU had the same weight prior to interviewing. A random integer was selected between 0 and the within-PSU SI to determine the first HU on the list that was in sample. The SI was added to the random number to identify the next HU in sample. This process, adding the SI to the most recent HU and identifying the sampled HUs, continued until the list of HUs, within the PSU, was exhausted. This sampling process was implemented in all PSUs, both SR and NSR.


P a g e | 25

A total of 22,725 total housing units were selected into sample. These addresses were mailed a pre-screener sample form.

Subsampling of Pre-Screener Sample

The eligibility of the 22,725 sampled addresses to receive a full detailed questionnaire was based on the mail out and internet results of the pre-screener questionnaire. As described earlier, the plans were to include 8,000 cases in sample split between the CAPI and CATI operations. Addresses that responded indicating that someone in the household had or was planning to participate in fishing, hunting, or wildlife-watching activities throughout the year were eligible to receive a detailed screener interview. Households were put into five categories based on the response status or combinations of answers to the pre-screener questionnaire. These categories were an Undeliverable/Ineligible group, a Nonparticipant group, a CATI Certainty group, a CAPI Certainty group, and a CAPI Eligible group. These groups were identified as follows:

1. Questionnaires that were returned from the post office due to the addresses being nonexistent, an incomplete address from the MAF, vacant, commercial, or undeliverable for some reason were classified into the Undeliverable/Ineligible group. These addresses became out-of-scope for the FHWAR survey.

2. Questionnaires that were returned with sufficient data and indicated that no one in the household was likely to participate in any wildlife-related activities in 2016 were classified into the Nonparticipant group. These cases were considered good FHWAR responses and are included in the response rate calculations in Chapter 4.

3. Questionnaires returned with a valid phone number, sufficient demographic data, and indication that someone in the household was likely to participate in fishing, hunting or wildlife-watching activities in 2016 were classified into the CATI Certainty group.

4. Questionnaires returned without a valid phone number but containing sufficient data and indicating that someone in the household was likely to participate in fishing, hunting, or wildlife-watching activities in 2016 were classified into the CAPI Certainty group. These addresses were included in the detail screener operation with certainty.

5. All other questionnaires were classified into the CAPI Eligible group for subsampling.


P a g e | 26

Table 3.5 provides the results of the pre-screener mailing.

Table 3.5 Eligibility Results of the Pre-screener Mailing

Classified Collection Mode Eligible Group Frequency

Undeliverable/Ineligible 3,266 Nonparticipant 2,772 CATI Certainty 3,339 CAPI Certainty 268 CAPI Eligible 13,080

Total 22,725

The 2,772 cases reporting on the pre-screener that they were not going to participate in the designated activities in 2016 were considered a complete interview. No further follow-up was conducted at these addresses. The 3,339 cases classified into the CATI group were sent to the CATI facility in Jeffersonville, IN to collect the detailed screener by phone. The 268 CAPI certainty cases were sent to the field for data collection by personal visit. The survey budget allowed for a total of 4,000 CATI cases. Since fewer than 4,000 cases were in the CATI group, a phone number look-up operation took place to find a phone number associated with the address. A sample of 691 cases that had a secondary phone number, one not provided by the respondent, were selected for a CATI interview and sent to the CATI facility to collect the detailed screener by phone. The remaining CAPI eligible cases were eligible for a personal interview. A subsample of 3,732 cases were selected for a total CAPI workload of 4,000 cases. Table 3.6 provides the results of all the subsampling.

Table 3.6 Mode of Data Collection

Mode Frequency

CATI 3,33916 CATI/CAPI17 691 CAPI 4,000 Nonparticipant (Pre-screener) 2,772

Total 10,802

16Thirty (30) cases responded to the pre-screener indicating that no one was going to participate. However, they did not provide enough information to include them in the weighting. Therefore, the Census Bureau followed up with these cases but did not count them as part of the 4,000 CATI cases. 17A phone number was found through a phone number look up operation and a subsample of CAPI eligible cases were selected to be sent to CATI.


P a g e | 27

Chapter 4. DATA COLLECTION METHODS

INTRODUCTION

This chapter details the 2016 FHWAR data collection methods designed to obtain quality data

and optimize response rates within budget and scheduling constraints. The FHWAR collects

data for a calendar year. In order to reduce recall bias, the data were collected about every

four-eight months. For the 2016 FHWAR, the Census Bureau collected data via a pre-screener

web/paper questionnaire, followed by a screener interview. Based on likely participation

reported in the screener interview, household members were selected for detail interviews.

PRE-SCREENER INTERNET/PAPER QUIESTIONNAIRE

The pre-screener operation consisted of a web-based instrument with a supplemental paper

questionnaire, if a web response was not received from the sample address. The Census Bureau

introduced the pre-screener internet/paper questionnaire operation based on a test of the

process conducted in 2013. Results from the 2013 test and justification for incorporating that

test methodology into the 2016 FHWAR are detailed in Chapter 2 of this document.

The pre-screener operation was conducted from January 4, 2016 through February 19, 2016.

The first mailout was sent to all households on January 4, 2016 and consisted of the FHW-PS(L1)

letter (refer to Appendix A, page 72, for a copy), with an invitation to access the website with

the specified username and password. A 2011 QuickFacts brochure (refer to Appendix A, page

79, for a copy), was enclosed with each letter to illustrate the types of data collected in the

FHWAR.

On January 15, 2016, the Census Bureau sent the FHW-PS(L2) reminder letter (which was very

similar to the first letter) with a new username and password to each household where a

response had not been recorded. The username and password from the first mailing were still

valid but any entered data were not accessible with the username/password provided in the

reminder letter. Included in this mailing was a paper questionnaire and return envelope to

allow household respondents the flexibility to respond in a mode that was preferable to them

(refer to Appendix B, pages 88-91, for a copy of the paper questionnaire).

On January 29, 2016, the Census Bureau sent a final copy of the FHW-PS(L2) letter with a new

username and password to each household where a response still had not been recorded. The

username and password from the first and second mailings were still valid but any entered data

were not accessible with the username/password provided in the third letter. Included in this

mailing was another copy of the paper questionnaire and a return envelope.

In each mailing, a toll-free number was included for respondents to contact the Census Bureau

telephone interview staff with questions or concerns. The Census Bureau interviewers


P a g e | 28

attempted to conduct the interview with respondents who called in with questions. Data were

entered through an administrative module of the internet instrument.

The paper questionnaires were returned to the Census Bureau National Processing Center

(NPC) in Jeffersonville, Indiana. Once they arrived there, they underwent minor editing as well

as keying and were transmitted to the Census Bureau programmer for further editing and

processing. Data from internet submission, including data gathered by the Census Bureau

interviewers, were transmitted directly to the Census Bureau programmer. Internet interviews

were accepted through February 16, 2016. The Census Bureau shut down the internet site as of

February 17, 2016. Paper questionnaires were accepted through February 19, 2016 to allow for

final paper questionnaires to arrive at NPC. Chapter 5 of this document contains a detailed

description of the editing and processing of the data.

WAVE 1 SCREENER AND FIRST DETAIL INTERVIEW

Based on the responses to the pre-screener questionnaire, a Wave 1 sample was selected as

outlined in Chapter 2 of this document. The Wave 1 operation was conducted in two modes:

Computer-Assisted Telephone Interviewing (CATI) and Computer-Assisted Personal

Interviewing (CAPI). Cases with a viable phone number collected through the pre-screener or

through vendor research (described in detail in Chapter 5) were assigned to CATI and cases

without a phone number or nonresponsive to the pre-screener mailings were assigned to CAPI.

On March 23, 2016, all sample households were mailed the FHW-W1(L) advance letter, a 2011

QuickFacts brochure and FH-1 Reference Aid (refer to Appendix A, pages 73, 79, and 80-87,

respectively, for copies). The Reference Aid was a collection of response options from a variety

of questions in the questionnaire. The Aid was used to inform the household of the types of

questions they would be asked in the upcoming interview and to help remind them of activity

they participated in or purchases they may have made since January 1, 2016.

CATI and CAPI interviewing began on April 1, 2016 and continued through May 15, 2016 (refer

to Appendix B, pages 92-94, for an abridged list of the questions asked). Both the CATI and CAPI

instruments were designed to interview one respondent for a screener interview and the

responses covered all members of the household. This household-level interview included the

household roster, basic demographic information for each household member (age, sex,

marital status, Hispanic origin, race, etc.), household income, and whether or not the

household member had participated in fishing, hunting, or wildlife-related activities such as

observing, feeding, or photographing wildlife so far in 2016. The screener interview also

included questions on how likely current nonparticipants would be to participate in any of

these activities during the remainder of the year.


P a g e | 29

Based on the responses to the screener questionnaire, the instrument sampled household

members for a detail interview (refer to Appendix B, pages 95-126, for an abridged list of the

questions asked). Eligible household members 16 years and older who had hunted or fished

between January 1, 2016 and the date of the screener interview were selected for a

sportsperson detail interview. Approximately 20% of household members 16 years and older

who had observed, fed, or photographed wildlife between January 1, 2016 and the date of the

screener interview were selected for a wildlife-watching detail interview. Household members

could be selected for both the sportsperson and wildlife-watching sample – this is referred to as

the “combo” sample.

WAVE 2 CATI DETAIL INTERVIEW AND WAVE 2 CAPI SCREENER AND DETAIL INTERVIEW

Wave 2 CATI Detail Interview

Once the Wave 1 screener data were processed, the Census Bureau subsampled household

members for a Wave 2 sportsperson, wildlife-watching participant, or combination detail

interview, based on criteria outlined in Chapter 3 of this document. If a household member was

selected for a detail interview in Wave 1, they could not be selected for a Wave 2 interview for

that sample (sportsperson or wildlife-watching). It was possible that a household member could

be selected for a detail interview in one of the samples in Wave 1 and then be selected for a

detail interview in the other sample for Wave 2. In 2016, 83.6% of the sample was selected for

both samples.

The instrument was updated to conduct a person-based interview for Wave 2. Each household

member selected for a Wave 2 interview was a unique case. Therefore, the Wave 2 workload

included multi-unit households where a household had more than one case assigned to it. The

Census Bureau systems were adjusted to accommodate linking multi-unit cases so that the

interviewer could ask for all appropriate respondents with one call attempt. Wave 2 was

conducted in CATI only. The questions and flow of the interview in the Wave 2 person-based

instrument were identical to the Wave 1 detail instrument questions and interview flow.

On August 25, 2016, household members selected for a Wave 2 interview were mailed the

FHW-W2(L) advance letter (refer to Appendix A, page 74, for a copy) with an FH-1 Reference

Aid. The Reference Aid was the same document that was mailed in the Wave 1 mailing. The

Census Bureau CATI interviewers began contacting respondents on September 1, 2016 by

telephone. Interviewing concluded on October 16, 2016. If a respondent had moved or changed

phone numbers since the Wave 1 interview, the interviewer was instructed to collect the new

address and phone number and attempt the case at the new contact number.


P a g e | 30

Wave 2 CAPI Screener Interview

During Wave 1, the Census Bureau CATI interviewers experienced difficulty reaching

households for a fairly large portion of the Wave 1 CATI sample. To decrease the nonresponse

rate, the Census Bureau incorporated a Wave 2 CAPI screener interview into the methodology

and sent a subsample of the Wave 1 CATI nonresponse cases for a CAPI personal visit. To

accomplish this additional operation, the CAPI Wave 1 instrument was updated to allow for

additional months of activities. Materials used in Wave 1 were provided to the field

representatives (FRs), along with the notes from the unsuccessful Wave 1 CATI interview

attempt, to increase the potential of a successful personal interview. As in Wave 1, households

were mailed the FHW-W1(L) advance letter with the 2011 QuickFacts brochure and the FH-1

Reference Aid. The Wave 2 CAPI screener operation ran concurrent with the Wave 2 CATI detail

interview (September 1, 2016 through October 16, 2016).

WAVE 3 CATI AND CAPI DETAIL INTERVIEW

The Wave 3 interview was paramount to the success of the 2016 FHWAR. If a respondent did

not complete a Wave 3 interview, the case was considered a non-interview and all prior detail

interview responses could not be used to produce estimates, since a full year’s worth of data

had not been collected. Due to the importance of this last wave of interviewing, the Census

Bureau scheduled a longer interview period and authorized more attempts to get a complete

interview.

All detail sample persons from Wave 1 and Wave 2 were included for a Wave 3 interview. If a

respondent had been selected for a sportsperson or wildlife-watching interview in Wave 1 and

then selected for the other interview (wildlife-watching or sportsperson, respectively) for Wave

2, the respondent was included in both samples for the Wave 3 interview. As a reminder, 83.6%

of the sample was selected for both samples in the 2016 survey.

Data from Waves 1 and 2 were provided in the Wave 3 input file to remind the respondent of

prior responses and reduce double-reporting of participation and expenditures. In addition,

questions regarding licenses and tags, land leasing and ownership, and big-ticket items, such as

a motor boat purchase, were added to this wave of interviewing (refer to Appendix B, pages 95-

126, for an abridged list of questions asked in Wave 3).

The Wave 3 interview was conducted in both CATI and CAPI. As in Wave 2, the Wave 3

workload included multi-unit households where a household had more than one case assigned

to it. The Census Bureau CATI system accommodated linking multi-unit cases so that the

interviewer could ask for all appropriate respondents with one call attempt. The CAPI

interviewers grouped their cases by address so they could complete multi-unit cases in minimal

visits or follow-up phone calls.


P a g e | 31

As in Wave 2, if the CATI or CAPI interviewer determined that the respondent had moved, they

attempted to get the new address and phone number for the respondent and attempted to

contact the respondent at the new number or address.

On December 21, 2016, all Wave 3 respondents were mailed one of three advance letters.

Cases where an initial detail interview had not been completed were mailed an FHW-W3 NI(L),

CAPI cases where an initial interview was completed were mailed the FHW-W3(L2) advance

letter, and CATI cases where an initial interview was completed were mailed the FHW-W3(L1)

advance letter (refer to Appendix A, pages 75-77, for copies). The FHW-W3 NI(L) letter

explained that the household was contacted earlier in the year and stressed the importance of

the survey. The FHW-W3(L2) and FHW-W3(L1) advance letters thanked the respondent for their

previous participation and stressed that this would be the last interview for the survey. The

FHW-W3(L1) advance letter also included a toll-free number for respondents to contact the

telephone interviewer to complete their interview. Each advance letter included the FH-1.3

Reference Aid which was the similar to the Reference Aid mailed in Waves 1 and 2. The FH-1.3

Reference Aid included some additional collections of response options due to the addition of

large equipment purchase questions to the Wave 3 instrument. Similar to the FH-1 Reference

Aid, this Aid was used to inform the respondent of the types of questions they would be asked

in the upcoming interview and to help remind them of activity they participated in or purchases

they may have made since their last interview.

The Census Bureau interviewers began contacting respondents on January 3, 2017 by

telephone and personal visit. The Census Bureau CATI interviewers attempted each CATI case at

least once through January 22. On January 23, the Census Bureau programmer and survey

methodologist subsampled CATI cases that had not had any contact. These subsampled cases

were recycled to the CAPI operation for a personal interview. The CATI interviewers continued

to attempt the remaining CATI cases through closeout. CATI and CAPI interviewing concluded

on February 28, 2017.

SUMMARY OF COMPLETE AND SUFFICIENT PARTIALS BY MODE OF INTERVIEW AND NUMBER OF CONTACTS

Table 4.1 describes the criteria for defining surveys as complete. Tables 4.2 through 4.5 detail

the number of cases completed by mode of interview and number of contact attempts per

case.

Each questionnaire had a separate criteria for determining a sufficient partial or complete. The

pre-screener was the only questionnaire a respondent was given an option of completing on

paper (versus one done on a computer, even if the respondent did a phone interview). As such,

for the pre-screener at the minimum a specific section had to be completely filled out to be

classified as a sufficient partial. In the computer assisted screener/detail sportsperson/detail


P a g e | 32

wildlife-watching questionnaires, since the questions had to be answered in the order the

computer dictated a specific question had to be reached/answered (even if the answer was a

refusal/don’t know response).

Table 4.1 Criteria for Defining Returned Surveys as Complete

Sufficient Partial Complete

Pre-Screener All questions in the demographic section (Step 5 – household composition) completed OR all questions in the avidity section (Step 6 – did anyone in household hunt/fish/wildlife-watch) completed.

All sections, including providing a phone number, were answered (Don’t Know (D) and Refused (R) were considered valid responses).

Screener N/A – stage.

No sufficient partials at this The household income question contained a response, which was the last survey question asked of the household respondent (D and R were considered valid responses).

Wave Surveys

Detailed Sportsperson

(includes fishing and hunting)

Completed questions up to and through the yes/no questions for the equipment primarily used for fishing (F_EQP2_J – “any other purchases” - which signified 80% of the survey had been completed) or hunting equipment if the respondent did not fish.

Completed the FH_OBSRV question (did sportsperson respondent participate in wildlife-watching activities at least 1 mile from home since last interview? – D and R were considered valid responses). This question came after all fishing and hunting questions; thus, all fishing and hunting questions were administered and answered.

Wildlife Completed the yes/no questions for the wildlife-watching equipment purchases (NCUEQP2_H – “any other purchases” - which signified at least 80% of the survey had been completed).

Completed NCU_FISH (did wildlife-watching respondent participate in fishing activities?).

Combination (sportsperson &

wildlife watching)

Completed the yes/no questions for the wildlife-watching equipment purchases (NCUEQP2_H – “any other purchases” - which signified at least 80% of the survey had been completed).

The interviewer reached the THANKYOU field. Because the combination cases did not receive the NCU_FISH or FH_OBSRV crossover questions.

Table 4.2 details screener cases by mode. The screener interview did not have a sufficient

partial classification so values are reported for completed cases only.


P a g e | 33

Table 4.2 Summary of Complete Screener Cases by Mode of Interview

Mode of Interview Screener Cases CATI CAPI

Wave 1 – Complete 2,241 2,641 Wave 2 – Complete N/A18 1,232

Total 2,241 3,873

Table 4.3 details the number of complete and sufficient partial sportsperson and wildlife-

watching cases by mode. Note that respondents could be in both the sportsperson and wildlife-

watching samples. Thus, it is not possible to sum the sportsperson total and wildlife-watching

total to calculate the total respondents as there would be double-counting of respondents.

Table 4.3 Summary of Complete and Sufficient Partial Detail Cases by Mode of Interview

Mode of Interview

WA

VE

3

WA

VE

2

WA

VE

1

Detail Cases CATI

Sportsperson, Total 371

Complete Interview 360

Sufficient Partial Interview 11

Wildlife Watching, Total 296

Complete Interview 293


Wave 1, Total19 629

Sportsperson, Total 1,840 Complete Interview 1,830


Wildlife Watching, Total 1,933

Complete Interview 1,923


Wave 2, Total 2,004

Sportsperson, Total 1,370



Wildlife Watching, Total 1,509



Wave 3, Total 1,598

CAPI

259

259

0

181

180

1

414

371 369

2

94

94

0

440

2,579

2,570

9

2,509

2,500

9

2,741

18Wave 2 Screener was CAPI only. 19Note that the wave totals are not the sum of the sportsperson total and wildlife-watching total because a respondent could be in both the sportsperson and wildlife-watching samples. Rather, these totals represent the number of respondents in each wave

by mode of interview.


P a g e | 34

Tables 4.4 and 4.5 summarize the number of contacts for complete or sufficient partial cases by

interview period. Table 4.4 summarizes the number of contacts for complete and sufficient

partial cases for household-level cases (pre-screener and screener). Note that the pre-screener

paper questionnaire had a maximum of two contacts as a paper questionnaire was only

included in the second and third mailings.

Table 4.4 Number of Contacts for Complete or Sufficient Partial Household-level Cases by Wave and Mode of Interview

Number of Contacts

Pre-Screener 1 2 3

Internet Complete

Sufficient Partial

Interview

Interview

694

26

736

32

782

36

Paper Complete

Sufficient Partial

Interview

Interview

3,455

542

78

1

N/A

N/A

20Screener 1-5

6-10 11+

Wave 1 CATI Wave 1 CAPI Wave 2 CAPI21

1,537 1,695 940

442 463 219

262 120 72

Table 4.5 summarizes the number of contacts for complete and sufficient partial cases for

person-level cases (Wave 2 CATI, Wave 3 CATI, and Wave 3 CAPI). There were 363 Wave 1 CAPI

and one Wave 2 CAPI households that had no contact attempt information. There were 12

Wave 3 CAPI cases that had no contact attempt information.

20There were no sufficient partials in the Screener interview, therefore the number of contact attempts are for complete interviews. 21The Wave 2 screener workload was a subset of cases which were unsuccessfully attempted in Wave 1 CATI. The values in the Wave 2 screener rows indicate the number of contacts made during Wave 2 only and do not include any contact attempts made during Wave 1.


P a g e | 35

Table 4.5 Number of Contacts for Completed or Sufficient Partial Detail Cases by Mode of Interview

Number 22of Contacts

Person-Level Cases 1-5 6-10 11+ W

AV

E 2

CA

TI Single-Unit

Complete Sufficient

Cases, Total Interview Partial Interview

488 485

3

118 116

2

64 62 2

Multi-Unit Cases, Total Complete Interview Sufficient Partial Interview

619 617

2

582 580

2

133 132

1

WA

VE

3 C

ATI

Single-Unit

Complete

Sufficient

Cases, Total

Interview

Partial Interview

485

482

3

82

82

0

38

38

0

Multi-Unit Cases, Total

Complete Interview

Sufficient Partial Interview

599

596

3

227

224

3

167

164

3

WA

VE

3 C

AP

I Originated in CAPI, Total

Complete Interview


1,638

1,634

4

361

358

3

62

61

1

Recycled from CATI, Total

Complete Interview


571

570

1

78

78

0

19

19

0

RESPONSE RATES

Several response rates for the National Survey of Fishing, Hunting, and Wildlife-Associated

Recreation can be calculated. Presented here are the response rates for the Wave 1 screener

operations, Wave 2 screener operations, and Wave 3 operations. The Wave 1 and 2 screener

operations are household-level response rates while the Wave 3 operation was based on

person-level response rate.

22For the Wave 2 Multi-Unit, Wave 3 Multi-Unit and Wave 3 CAPI (both Originated and Recycled) cases, the amount of contacts

was averaged between household members in sample. The total number of contacts recorded for each household member in

sample was combined and divided by the total number of household members in sample. That calculated sum (rounded to a

whole number) was then set as the amount of contacts for each sample person in that household.


P a g e | 36

AAPOR Response Rate Calculator

Calculations for the 2016 FHWAR response rates were based on the AAPOR response rate

calculator (Version 3.1, created in November 2010) found on the AAPOR website.23 The CATI

and CAPI outcome codes were mapped to the closest definable rows in the response rate

calculator.

National Screener Response Rates

Table 4.6 provides the AAPOR response rate calculator with the results of the household data

collections for both Wave 1 and Wave 2. Two columns are provided for each wave for the

different modes of data collection. The Wave 2 columns combine all screener interviews for

both screener waves. The cases that were considered recycled from Wave 1 had the outcome

code overwritten with the Wave 2 CAPI outcome.24,25 These counts are included in the Wave 2

CAPI outcome column. The last column of the calculator is the sum of the CATI and CAPI

operations that include the 2,772 pre-screener households that said that no one in the

household was going to participate in any type of FHWAR activities.

Table 4.6 AAPOR Outcome Rate Calculator -- Household Screener Response Rates

Wave 1 Wave 2 Final CATI CAPI CATI CAPI

Total phone numbers used 4030 4000 2333 5697 10802

Complete Interviews (I) 2103 2630 2103 3842 8717

Partial Interviews (P) 138 11 138 31 169

Refusal and break off (R) 565 528 79 778 857

Non-Contact (NC) 990 162 0 220 220

Other (O) 211 0 13 0 13

Unknown Household (UH) 0 0 0 0 0

Unknown other (UO) 0 0 0 0 0

Estimated proportion of cases of unknown eligibility that are eligible (e)26

0.994 0.833 1.000 0.855 0.924

Response Rates Response Rate 1: I/(I+P) + (R+NC+O) + (UH+UO) 0.525 0.790 0.901 0.789 0.874 Response Rate 2: (I+P)/(I+P) + (R+NC+O) + (UH+UO) 0.559 0.793 0.961 0.795 0.891 Response Rate 3: I/((I+P) + (R+NC+O) + e(UH+UO) ) 0.525 0.790 0.901 0.789 0.874 Response Rate 4: (I+P)/((I+P) + (R+NC+O) + e(UH+UO)) 0.559 0.793 0.961 0.795 0.891

Cooperation Rates Cooperation Rate 1: I/(I+P)+R+O) 0.697 0.830 0.901 0.826 0.894

23The AAPOR website is www.aapor.org, accessed on May 10, 2016. 24CATI recycled cases switching mode of data collection from CATI to CAPI included outcome codes 020, 181, 183, 188, 193, 194, and 195 from the Wave 1 CATI operation. 25See Appendix E on pages 141-142 for a list of CATI and CAPI outcome codes. 26Enter a different value or accept the estimate in this line as a default. This estimate is based on the proportion of eligible units among all units in the sample for which a definitive determination of status was obtained (a conservative estimate). This will be used if you do not enter a different estimate. For guidance about how to compute other estimates of e, see AAPOR's 2009 Eligibility Estimates.

http://www.aapor.org/


P a g e | 37

Wave 1 Wave 2 Final CATI CAPI CATI CAPI

Cooperation Rate 2: (I+P)/((I+P)+R+0)) 0.743 0.833 0.961 0.833 0.911 Cooperation Rate 3: I/((I+P)+R)) 0.749 0.830 0.906 0.826 0.895 Cooperation Rate 4: (I+P)/((I+P)+R)) 0.799 0.833 0.966 0.833 0.912

Refusal Rates Refusal Rate 1: R/((I+P)+(R+NC+O) + UH + UO)) 0.141 0.159 0.034 0.160 0.086 Refusal Rate 2: R/((I+P)+(R+NC+O) + e(UH + UO)) 0.141 0.159 0.034 0.160 0.086 Refusal Rate 3: R/((I+P)+(R+NC+O)) 0.141 0.159 0.034 0.160 0.086

Contact Rates Contact Rate 1: (I+P)+R+O / (I+P)+R+O+NC+ (UH + UO) 0.753 0.951 1.000 0.955 0.978 Contact Rate 2: (I+P)+R+O / (I+P)+R+O+NC + e(UH+UO) 0.753 0.951 1.000 0.955 0.978 Contact Rate 3: (I+P)+R+O / (I+P)+R+O+NC 0.753 0.951 1.000 0.955 0.978

Final Wave 3 Detail Person Interview Response Rates

Table 4.7 provides the AAPOR response rate calculator with the results of the person-level

detail questionnaire results from Wave 3. Two columns are provided for the two modes of data

collection. The last column of the calculator is the sum of the CATI and CAPI operations.

Table 4.7 AAPOR Outcome Rate Calculator -- Wave 3 Detail Person Interview Response Rates

CATI CAPI Total

Total phone numbers used Complete Interviews (I) Partial Interviews (P) Refusal and break off (R)

2907 1586

12 521

3797 2732

9 572

6704 4318

21 1093

Non-Contact (NC) Other (O) Unknown Household (UH) Unknown other (UO) Estimated proportion of cases of unknown eligibility that are eligible (e)27

532 178

0.973

401

0.978

933 178

0 0

0.976

Response Rates Response Rate 1: I/(I+P) + (R+NC+O) + (UH+UO) 0.561 0.736 0.660

Response Rate 2: (I+P)/(I+P) + (R+NC+O) + (UH+UO) 0.565 0.738 0.663

Response Rate 3: I/((I+P) + (R+NC+O) + e(UH+UO) ) Response Rate 4: (I+P)/((I+P) + (R+NC+O) + e(UH+UO))

0.561 0.565

0.736 0.738

0.660 0.663

Cooperation Rates Cooperation Rate 1: I/(I+P)+R+O) Cooperation Rate 2: (I+P)/((I+P)+R+0)) Cooperation Rate 3: I/((I+P)+R)) Cooperation Rate 4: (I+P)/((I+P)+R))

0.690 0.696 0.748 0.754

0.825 0.827 0.825 0.827

0.770 0.773 0.795 0.799

Refusal Rates

27Enter a different value or accept the estimate in this line as a default. This estimate is based on the proportion of eligible units among all units in the sample for which a definitive determination of status was obtained (a conservative estimate). This will be used if you do not enter a different estimate. For guidance about how to compute other estimates of e, see AAPOR's 2009 Eligibility Estimates.


P a g e | 38

CATI CAPI Total Refusal Rate 1: R/((I+P)+(R+NC+O) + UH + UO)) 0.184 0.154 0.167 Refusal Rate 2: R/((I+P)+(R+NC+O) + e(UH + UO)) 0.184 0.154 0.167 Refusal Rate 3: R/((I+P)+(R+NC+O)) 0.184 0.154 0.167

Contact Rates Contact Rate 1: (I+P)+R+O / (I+P)+R+O+NC+ (UH + UO) 0.812 0.892 0.857 Contact Rate 2: (I+P)+R+O / (I+P)+R+O+NC + e(UH+UO) 0.812 0.892 0.857 Contact Rate 3: (I+P)+R+O / (I+P)+R+O+NC 0.812 0.892 0.857

Data Preparation and Processing 2016 FHWAR Design and Methodology Technical Report

P a g e | 39

Chapter 5. DATA PREPARATION AND PROCESSING

INTRODUCTION

The main purpose of data preparation and processing is to take the response data gathered

from each survey collection mode and survey wave to the point where it can be used to

produce survey estimates. Data returned from the field and telephone center typically arrive in

various stages of completion, from a completed interview with no problems to one with most

or all of the data items left blank. There can be inconsistencies within the interviews, such that

one response contradicts another.

Many processing procedures were necessary to prepare the 2016 FHWAR data for tabulation.

This chapter details each data preparation procedure separately. There were procedures that

occurred before each interview period, after the pre-screener, and after each of the three main

waves of interviewing. Additional procedures were executed once all interviewing was

completed. The final products produced were statistical tables and publicly released data text

files (with codebooks and SAS conversion programs).

PREPARATION TO CREATE LABEL FILES AND INPUT FILES

Address Standardization

The initial sample was selected from MAF. To identify blank or incomplete addresses that

should not be selected for use in the FHWAR survey, the sample was run through the NPC

address standardization process. The process identified addresses that were deemed

nonexistent/unmailable and they were removed from the sample. The remaining addresses

were standardized to all have consistent formatting (i.e., placement of directional character in

street name, similar spelling/abbreviations for street endings).

Telephone Research

Since the MAF did not contain phone numbers, the Census Bureau conducted a phone number

lookup operation before the Wave 1 data collection operation. The MAFID for each sample case

was sent to the Census Bureau’s Center for Administrative Records Research and Applications

(CARRA) where a phone number matching that location was returned if available. If a phone

number was provided by a respondent in the pre-screener operation and the household was

selected for Wave 1, that phone number was used as the primary contact number and any

researched phone number was used as a secondary contact. If no phone number was returned

in the pre-screener operation, the researched phone number, if available, was used as primary

contact.

Input Files

Input files providing information for the instrument during interviewing were created for each

wave. Wave 1 and Wave 2 screener input files were household-based and contained


P a g e | 40

information including case identification number, household address, phone number (if

present), and some geographic information (if present, helpful to find an address if the case

was assigned for a personal visit). The Wave 2 CATI input file included the same information as

the Wave 1 and Wave 2 screener files, plus the respondent name/age/sex, any previous

interviewer notes, and contact information (to aid in locating the respondent in a future

interview). The Wave 3 input file included everything in the Wave 2 CATI input file plus select

dependent data collected from previous Wave 1 and Wave 2 interviews to aid in recall of

previous activity and purchases to avoid double-reporting of participation and expenditures.

DATA PROCESSING PER WAVE OF INTERVIEWING

The 2016 FHWAR had four set interviewing periods. The pre-screener, used to collect a phone

number for the household and basic household composition, was conducted between January

4, 2016 and February 19, 2016. Wave 1 interviewing took place April 1, 2016 through May 15,

2016. Wave 2 interviewing took place September 1, 2016 through October 16, 2016. Wave 3

interviewing took place January 3, 2017 through February 28, 2017.

Pre-screener Interview Data Processing

The pre-screener interview was conducted both by paper questionnaire and internet response.

Paper questionnaires were processed and keyed at NPC, with data transmitted back to the

Census Bureau programmer. Consistency edits were performed on both paper and internet

responses. Invalid phone numbers (i.e., all same digits, such as 333-333-3333 and obvious fake

phone numbers, such as 123-456-7890) were blanked out.

Responses for the household composition section (total number of household members,

number of people per age/sex group) were compared for consistency. If the total household

member count was blank but individual member counts were provided, the summed member

count was stored in the household count variable. If the household member count deviated

from individual/summed member counts or all member counts were blank, individual member

counts were adjusted.

An overall outcome status code was created for the pre-screener based on presence of phone

number(s), presence of data in the household roster section, and presence of data in the basic

avidity sections (did anyone in household hunt/fish/wildlife watch). The pre-screener outcome

code was used to determine sample eligibility for a Wave 1 interview.

Screener Interview Data Processing

One household respondent provided the screener data for each member of the household.

Data collected for the majority of questions were stored in arrays at the household level. Each

member of the household was assigned a line number in the household roster. Post data-

collection editing recoded the responses into person-level records. Fields where a household

member’s line number was reported for various avidity questions were recoded to reflect the


P a g e | 41

answer to the lead-in question (i.e., “Did you/anyone in the household hunt last year?”), so the

line number response would be recoded to “1” (for yes) for that specific person record. Wave 2

(CATI) and Wave 3 interview operations were person-based so no data needed to be recoded in

that manner.

If a person usually resided at an alternative address, they were deleted from the household

roster and were ineligible for any further screener questions or sampling into a detail interview.

Detail Interview Data Processing

Detail interviews for all three waves generated similar output files from the computer

instrument, with Wave 3 having additional files for annual expenditures and license/cost data.

Each section of the detail interview instrument that had rostered data (i.e., states hunted in and

related information, types of game hunted in each state) generated a separate output file. The

end result of processing for each wave was a single combined data file, with one record per

person where rostered information was collapsed in arrays.

Shifting Responses

Certain questions in the instrument were multiple choice entries, where the numbered answer

categories were initially stored in the order they were provided by the respondent, even if that

order was different than displayed in the instrument question. To aid in the use of the public

data, each respondent answer choice was recoded to the value “1” for the specific answer

category (i.e., answer choice “4” was recoded to “1” for slot/array position four of that

variable/question name). This shifting of answers happened for the following interview

questions:

Which Great Lakes fished in (WHCHGL);

Which type of fish (Saltwater, Freshwater, Great Lakes) fished for, with each type collected

in a separate question in the instrument (SLTYP, FRTYP, GLTYP);

Did respondent observe/photograph/feed wildlife in trips in/to state (FH_OBSRV – national

question, for states: TRP1_Q1 - observe, TRP1_Q2 - photograph, TRP1_Q3 - feed);

Did the birds the respondent (observed/photographed/fed) in state include birds of prey,

waterfowl, other water birds, songbirds, and/or other birds (TYPBRD1 - birds of prey,

TYPBRD2 - waterfowl, TYPBRD3 - other water birds, TYPBRD4 - songbirds, TYPBRD5 - other

birds); and

Did the respondent on trips to state observe/photograph/feed fish, large land mammals,

small land mammals, marine mammals, and/or other wildlife (ANIMLS1 - fish, ANIMLS2 -

large land mammals, ANIMLS3 - small land mammals, ANIMLS4 - marine mammals,

ANIMLS5 - other wildlife).


P a g e | 42

The types of game hunted were stored both in the original order given by the respondent for

the original variable/question name, and also recoded for ease of public data use to a value of

“1” for specific slot/array positions in newly created variable names. All four types of game

hunted (big game, small game, migratory birds, and other animals) were collected in one

instrument question. For each type of game, a separate group of variables (named for the type

of game) were created where a “1” was stored if that specific game type was hunted.

Recoding “Other Specify” for Species

In addition to providing a set list of hunting/fishing species a respondent could choose from,

the instrument also provided “other specify” choices in which a text answer could be filled in.

Using an automated process, if that response exactly matched the spelling of any of the set

species already listed, the “other specify” response was blanked, the matched set species was

filled, and any information tied to the other specify (i.e., days hunted, trips taken to hunt other

specify species) was reallocated. In the final sportsperson public use data file only the fact that

the respondent indicated an “other specify” species is provided. The exact text response was

not provided for confidentiality reasons.

Filling Skipped Questions

If a question’s response could be exactly derived from other responses the respondent

provided, that question was skipped during interviewing and filled in automatically during data

processing. For example, respondents who fished in freshwater only states, or fished only for

freshwater fish in other states, were not asked for number of freshwater days in that state. The

value was taken from the respondent’s answer to the number of days fished in state overall or,

if the respondent only fished in one state, the number of days fished in the United States.

Backfilling Day Values

The survey asked multiple questions related to how many days a respondent did

hunting/fishing/wildlife-related activities. The days questions became more detailed as the

interview progressed into the specific areas of participation, such as big game hunting. To

ensure consistency among all answers in a direct logical chain, the more specific value was

compared to the more general one. Preference was always given to the more specific values

over the general values. If the more specific value was greater than the general value, the

general value was changed to the specific value. If the sum of a group of more specific day

values was less than a more general value and all specific day values had an entry (not including

Don’t know and Refused), the general value was reduced to the calculated sum.

Renaming Variables

Each wave produced a number of output text files. In order to combine all those files into one

main data file (one record per person), many of the variable names were renamed into arrays.

For each sample (sportsperson, wildlife watching) there could be up to two interviews for the

entire survey period. In most cases an “a” (for first interview) or “b” (for the second interview)


P a g e | 43

were added to the variable name, along with an array number whose maximum value

represented the amount of responses from the most avid respondent receiving that question.

Blanking Don’t Know/Refused Responses

Don’t Know (D) and Refused (R) responses were included in the instrument to allow the

respondent to bypass a question if they were uncertain of the response or were unwilling to

give one. Those responses were blanked out in data processing.

DATA PROCESSING AFTER THREE WAVES OF INTERVIEWING

After all three waves of interviewing were completed, additional processing occurred to

combine all the information collected in the three waves into one master record per person for

the entire survey. The objective was to have a consolidated file that was easy to use in creating

the statistical tables to be included in the National Report.

Renaming Variables

Variables that did not need to be renamed after each wave needed to be renamed when

merging the three waves to create a consolidated file. High-level variables such as whether or

not a respondent hunted in the United States in 2016 were given a “_1” or “_2” extension

(indicating the data were collected in either the first or second interview, the second interview

always referring to Wave 3). The final variable on the consolidated file maintained the original

variable name. For yes/no questions, if the respondent answered “1” (for yes) in either wave,

the final answer was “1”. If the respondent answered “2” (for no) in either wave and the other

wave’s response (if present) was “2”,”D” (for don’t know), or “R” (for refused), the final answer

was “2”. Day values for identical variables in each wave were summed, with the resulting value

limited to the maximum days in 2016, which was 366 as it was a leap year.

For variables previously renamed by adding an “a” or “b” (for interview wave), final versions of

those variables were created with a “d” extension of their original instrument name. All

variables of that type were arrays of variables so also have an array number after the added

letter ending. Information was copied into the “d” variables from the data for the first

successful interview. If there was a second successful interview, corresponding matching

variables (usually based on the state activity occurred in) were combined into one “d” entry.

New “d” array positions were used to store data from the second interview that could not be

merged with data from the first interview.

Some data in each interview wave had to be kept separate. For example, if a respondent

bought equipment items in the same category in both interviews, all associated variables (i.e.,

cost, primary use of equipment, state(s) equipment bought in) were stored in variable names

identified by the interview it was collected in. This was to allow the cost of the equipment to be

divided correctly by the number of state(s) entered for that interview wave alone. These

variables kept the names they were renamed to in each of the interview waves (a or b, 1 or 2).


P a g e | 44

Converting Percentages

For interview questions where the respondent was asked how much of their total trip cost to a

particular state was spent in their resident state, the instrument gave the option of either

providing a percentage or an exact dollar amount. During merging of the data for each wave,

any percentage figure given was converted to the equivalent dollar amount.

Recoding 1/2 (Yes/No) to 1/0

Researchers using the public use data in the past had expressed the desire to easily tabulate

how many positive answers a group of questions contained. “No” responses were changed

from 2 to 0 to allow the researcher to simply sum up all the variable values to achieve their

result, instead of having to individually test for “1” in each variable.

EDITING AND IMPUTATION

Once the data were combined, selected screener variables, if blank, were imputed. Those

variables were age, race, sex, relationship to household respondent, marital status, and

maximum schooling achieved (AGE, SCRACE, SEX, RELATION, MARITAL, and SCHOOL). The

imputation process was a mixture of relational imputation (infers missing value from other

persons/household characteristics) and “hot deck” allocation (assigns missing value from a

record with similar characteristics). Age was imputed based on spousal relationship,

parental/child relationship, or relationship to other household members, or it was assigned a

hot deck value based on other households in related geographic areas. Race was imputed

based on parental/child relationship, spousal relationship, relationship to other household

members, or assigned a hot deck value based on other households in related geographic areas.

Sex was imputed based on spousal relationship, or assigned to keep imputed gender counts

equal. Relationship to household respondent was imputed based on presence of a spousal

relationship in the household and age of person missing household relationship. Marital status

was imputed if there was a spouse indicated in the household relationship question, presence

of parents in household, or a household member was less than sixteen years old. Schooling was

only imputed for people aged 16 years and older based on schooling level of other household

members or assigned a hot deck value based on other households in related geographic areas.

Further imputation was conducted on demographic variables of the non-participant

respondents from the pre-screener questionnaire. Age, race, sex, Hispanic origin, education,

and household income were imputed for these cases since the collection of these data were

not included on the pre-screener form and further contact to non-participating respondents did

not occur. Therefore, distribution of these variables at the national level for all persons would

not have been achievable. Thus, the imputation looked at the distribution of the detailed

demographic variables and distributed the pre-screener respondents to the demographic

groups to match the distribution for these variables to the ACS distribution at the national level.


P a g e | 45

After the imputation process, the screener, sportsperson, and wildlife-watching weights were

calculated based on specifications from the Census Bureau’s Demographic Statistical Methods

Division (DSMD). The weights were used in creating the statistical tables and other values

reported in the National Report. Weighting and estimation are described in more detail in

Chapter 6 of this document.

Using the weighted master data file, the FWS reviewed special equipment expenditures and

land leasing/ownership values to identify outliers at the national level. Only those values were

examined due to their big impact on total expenditures. This step took place before topcoding

was implemented. Even though identifying outliers was a manual process, there was a specific

process/criteria on identifying values to be blanked.

Observations that had cost values above $100,000 were examined. For those

cases/respondents, records were flagged for further review if the household income was less

than the individual cost/purchase, if the respondent had other big purchases, or if the other

household members had high reported spending (especially if they reported identical values). A

value was deleted if the weighted expenditure was flagrantly outside the normal range of

values or the weighted observation made up a majority of the state’s total expenditures (more

common when a person from a big weight state bought something in a small weight state). A

total of eighteen outlier values were identified for the 2016 survey year.

The Census Bureau’s mandate to ensure confidentiality of released public data required a

process to topcode select types of variables. For those variables, the top three entries, or the

top three percent of entries if more than 100 entries in total, were averaged together. Since

questions in the 2016 FHWAR survey asking for the number of days for any activity had a

natural upper bound of 366 (since 2016 was a leap year) they were exempted from being

topcoded. The majority of values topcoded were expenditure values and some land leasing and

ownership values (i.e., number of acres, amount of members in group owning/leasing land).

Variances were created for select variables for the screener, sportsperson, and wildlife-

watching samples separately. Refer to Appendix C for a list of the variables in each sample

where direct variance values were calculated. Variances are covered in-depth in Chapter 7.

CREATING FINAL DATA PRODUCTS

To improve the usability of the public data, various summation variables were created. For

example, a variable was created to represent the total number of days a respondent hunted big

game in the United States in 2016. The value sums the entries for the instrument question

regarding number of big game hunting days by state, or, if no other type of game was hunted in

that state, the total number of hunting days for that state.


P a g e | 46

The 2016 FHWAR publicly-released data products were produced from a nationally-

representative sample28 unlike previous survey cycles which were both nationally and state

representative. The initial sample was drawn based on thirteen nationally-represented sample

areas (refer to Table 3.2 on page 17). Individual state values (i.e., state where the activity

occurred, state that the expenditure was attributed to) collected during the survey were

recoded to one of the first nine sample areas listed in Table 3.2. For the four test states where

the Census Bureau selected state-representative samples, the states were recoded as follows:

Maine was included with the “New England” sample,

Minnesota was included with the “West North Central” sample,

Oklahoma was included with the “West South Central” sample, and

Virginia was included with the “South Atlantic” sample.

To further guard against the possible identification of a single respondent, certain

variables/values released on the public use data files were suppressed. Some values of the

geographic variables (population density, population size) were suppressed if weighted counts

were less than a certain threshold when crossed with Census division. Even though a single race

category was selected for use in creating the statistical tables in the National Report, all five

multiple-choice race categories were provided on the public use file. For respondents that had

selected certain combinations of race categories, the five individual race categorical values

were suppressed, but the created single-race category/variable was kept in place. The data

suppression of geography and race was applied after the data presented in the National Report

was created, so not all values computed in the statistical tables can be recreated exactly using

the public use data.

From the master data file used for the National Report, three separate public use files were

created – a screener file, a file of sportsperson activity, and a file of wildlife-watching activity.

Certain screener variables were also included in the sportsperson/wildlife-watching files for

ease of data use. All three files were text files and SAS programs were created and released to

the public for use in converting the text files into SAS data sets. For each file, a codebook listing

all variable names, their sizes, start/stop locations on the flat file, and descriptions was created.

28The Census Bureau also selected state-representative sample for four test states: Maine, Minnesota, Oklahoma, and Virginia. However, the data from these states were for analysis purposes only and not publicly released.

Weighting and Estimation 2016 FHWAR Design and Methodology Technical Report

P a g e | 47

Chapter 6. WEIGHTING AND ESTIMATION

INTRODUCTION

A probability sample is defined as a sample that has a known non-zero probability of selection

for each sample unit. With probability samples, unbiased estimators can be obtained. These are

estimates that on average, over repeated samples, yield the population’s true values. An

unbiased estimator of the population total for any characteristic investigated in the survey may

be obtained by multiplying the value of that characteristic for each sample unit (person or

household) by the reciprocal of the probability with which that unit was selected and summing

the products over all units in the sample (Hansen, Hurwitz, and Madow, 1953). By starting with

unbiased estimates from a probability sample, various kinds of estimation and adjustment

procedures (such as for non-interview) can be applied with reasonable assurance that the

overall accuracy of the estimates will be improved. In the FHWAR sample, not all units

responded, and this nonresponse is a potential source of bias. This nonresponse rate was 10.9

percent29 for household units.

Other factors, such as occasional errors caused by the sample selection procedure or the

omission of households or individuals missed by interviewers, can also introduce bias. These

omitted households or people can be considered as having zero probability of selection. The

probability of selecting each unit in the FHWAR is known, and every attempt is made to keep

departures from true probability sampling to a minimum.

To produce FHWAR national estimates from survey data, a statistical weight for each person in

the sample was developed through the following steps. The first step created a screener weight

for all persons in the sampled households. The second step created a participation weight for

those who were eligible to answer the detail questionnaires. Two participation weights were

created, one for the sportspersons and the second for the wildlife-watching participants.

Creation of screener household weights involved the following steps:

1. Preparing a base weight derived from the FHWAR sampling probabilities; 2. Adjusting for subsampling of pre-screener units; 3. Adjusting for screener nonresponse; 4. Applying a first-stage ratio adjustment to reduce variances due to the sampling of PSUs; 5. Applying a second-stage ratio adjustment to reduce variances by controlling FHWAR

estimates of the population to independent estimates of the current population – to create the final screener weight.

29Using the American Association for Public Opinion Research (AAPOR) Response Rate calculator formula Response Rate 2. See Chapter 4.


P a g e | 48

Creation of the detailed sportsperson and wildlife-watching weights involved the following steps:

1. Applying an adjustment for detail interview nonresponse to the final screener weight; 2. Applying a ratio adjustment to screener person within defined participation strata.

Each of these steps is explained below.

SCREENER WEIGHTS

Base Weights

The sample designated in the FHWAR survey was selected with probabilities equal to the

inverse of the required region/state sampling intervals. These sampling intervals are called the

base weights. All sample households within the same region/state have the same probability of

selection. The base weight is assigned to every person in the sampled housing unit (HU).

Pre-screener Subsampling Factor

As described in Chapter 3, a subsampling operation was implemented to subsample the

addresses that did not respond to the pre-screener questionnaire. The pre-screener

subsampling factor adjusted the base weights of the case eligible for this subsampling

operation. The subsampling factor (SSF) was calculated as:

SSF= Number of eligible pre-screener nonrespondents

Desired number of CAPI interviews30

This factor was applied to only those addresses that did not respond to the pre-screener

questionnaire. All other cases received a pre-screener subsampling factor of 1.

The weight after this step is: (base weight) x (pre-screener subsampling factor)

Adjustment for Nonresponse

Nonresponse arises when households or other units of observation that have been selected for

inclusion in a survey fail to provide all or some of the data that were to be collected. This failure

to obtain complete results from all the units selected can arise from several different sources,

depending upon the survey situation. There are two major types of nonresponse: item

nonresponse and complete (or unit) nonresponse. Item nonresponse occurs when a

cooperating HU/person fails or refuses to provide some specific items of information.

Procedures for handling this type of nonresponse are discussed in Chapter 4. Unit nonresponse

refers to the failure to collect any survey data from an occupied sample HU. For example, data

30The desired number of CAPI interviews (4,000) in the denominator was reduce by the number of cases that responded in the pre-screener reporting that someone in the household was going to participate in an activity but did not provide a phone number to use in the CATI screener operation.


P a g e | 49

may not be obtained from an eligible household in the survey because of impassable roads, a

respondent’s absence or refusal to participate in the interview, or unavailability of the

respondent for other reasons.

In the FHWAR estimation process, the weights for all interviewed households are adjusted to

account for occupied sample households for which no information was obtained because of

unit nonresponse (Type A non-interviews). This non-interview adjustment is made separately

for four areas within each region/state. These areas are within:

1. The central city of a Metropolitan Statistical Area (MSA) 2. Balance of an MSA 3. Urban areas outside an MSA 4. Rural areas outside an MSA

The non-interview factor, NRFAj, is computed as:

𝑁𝑅𝐹𝐴𝑗 = 𝐼𝐴𝑗 + 𝑁𝑅𝐴𝑗

𝐼𝐴𝑗

where

Iij = the weighted count of interviewed households in cell j of region/state A, and

NRij = the weighted count of Type A non-interviewed households in cell j of region/state A.

At the completion of the non-interview adjustment procedure, the weight for each interviewed

person is: (base weight) x (pre-screener subsampling factor) x (non-interview adjustment

factor)

First-Stage Ratio Adjustment

The purpose of the first-stage ratio adjustment is to reduce the variance of region/state-level

estimates caused by the sampling of PSUs; that is, the variance that would still be associated

with the region/state-level estimates even if the survey included all households in every sample

PSU. This is called the between-PSU variance.

There are a couple of factors to consider in determining what information to use in applying the

first-stage adjustment. The information must be available for each PSU and correlated with as

many of the statistics of importance published from the FHWAR as possible. By using the

licensed hunter count, the first-stage ratio adjustment compensates for the fact that the

licensed hunter count composition of a NSR sample PSU could differ from the licensed hunter


P a g e | 50

count composition of the stratum it is representing. This adjustment is not necessary for SR

PSUs since they represent only themselves.

Computing First-Stage Ratio Adjustment Factors

The first-stage adjustment factors are based on hunter license information provided by the

states and are applied only to sample data for the NSR PSUs. Factors are computed for each

region/state containing NSR PSUs. The following formula was used to compute the first-stage

adjustment factors for each region/state:

𝐹𝑆𝐹𝐴 =∑ 𝐻𝑢𝑛𝑡𝑒𝑟 𝐶𝑜𝑢𝑛𝑡𝐴𝑖

𝑃𝑆𝑈𝐴𝑖=1

∑ 𝐻𝑢𝑛𝑡𝑒𝑟 𝐶𝑜𝑢𝑛𝑡𝐴𝑘 (1

𝜋𝐴𝑘)

𝑃𝑆𝑈𝑆𝑘=1

where

FSFA

= the first-stage factor for region/state

Hunter CountAi

=

the number of licensed hunters in NSR PSU i (sample or nonsample) in region/state A

Hunter CountAk

= the number of licensed hunters in NSR sample PSU k in region/state A

πAk

= probability of selection for sample PSU k in region/state A

PSUA

= total number of NSR PSUs (sample and nonsample) in region/state A

PSUS = number of sample NSR PSUs in region/state A The estimate in the denominator of each of the ratios is obtained by multiplying the number of

licensed hunters for each NSR sample PSU by the inverse of the probability of selection for that

PSU and summing over all NSR sample PSUs in the region/state.

At the completion of the first-stage ratio adjustment, the weight for each responding person is

the product of: (base weight) x (pre-screener subsampling factor) x (non-interview adjustment

factor) x (first-stage ratio adjustment factor).

The weight after the first-stage adjustment is called the first-stage weight.

Second-Stage Ratio Adjustment

The purpose of the second-stage factor is to ensure that the sample-based estimates of

population match independent population controls. The target population for the 2016 FHWAR

is the household population aged 16 years and older. However, prior year participation data are


P a g e | 51

collected for children aged 6 through 15 as reported in Appendix D of the National Report, and

so data were collected on the screener for all people aged 6 years old and older. Therefore, the

population controls for FHWAR were divided into two groups: those aged 6-15 and aged 16 and

older living in housing units.

For each region/state, the second-stage factor is:

𝑆𝑆𝐹𝐴𝑖 = 𝑃𝐴𝑖

𝑊𝐴𝑖

where

i = age group 6 to 15 years or age group 16 years and older A = is the region/state

PAi = the independent estimate of the population in region/state

WAi = the weighted survey estimate of the population in region/state. This is the sum of all first-stage person weights

At the completion of the second-stage ratio adjustment, the weight for each responding person

is the product of: (base weight) x (pre-screener subsampling factor) x (non-interview

adjustment factor) x (first-stage ratio adjustment factor) x (second-stage ratio adjustment

factor).

After this adjustment, this is the final screener weight. All persons within the sampled

households, regardless of participation status, receive a final screener weight.

DETAIL PERSON WEIGHTS

Two separate weights were created for the detail person weights. These weights are for the

two detail subject matter sections of the questionnaire: one for the sportsperson sample

questionnaire and the second for the wildlife-watching sample questionnaire. Details of each

are provided below.

Sportspersons Sample

Information provided in the screener questionnaire allows stratification of the household

members into strata based on their expected participation in sportspersons activities. For the

2016 FHWAR, the sportsperson strata were defined as:

If the person already fished or hunted in 2016, the person was a sportsperson and assigned to stratum 1. These people received a detail interview right after completing the screener questionnaire – also referred to as Wave 1 interviewing.


P a g e | 52

If the person last fished or hunted in 2015 for 10+ days or spent $100+, the person was defined as an avid sportsperson participant and assigned to stratum 2.

If the person last fished or hunted in 2015 for less than 10 days and spent less than $100, the person was defined as an average sportsperson participant and assigned to stratum 3.

If the person did not fish or hunt in 2015, but has fished or hunted since 2011 and was very likely to in 2016, the person was defined as an infrequent sportsperson participant and assigned to stratum 4.

If the person did not fish or hunt in 2015, but has fished or hunted since 2011 and was somewhat likely to in 2016, the person was defined as an inactive sportsperson participant and assigned to stratum 5.

If the person did not fish or hunt in 2015, but has fished or hunted since 2011 and was somewhat unlikely to in 2016, the person was defined as a nonparticipant and assigned to stratum 6.

If the person was very unlikely to fish or hunt in 2016 or has not fished or hunted since 2011, then the person was assigned to stratum 7. Households from the pre-screener that indicated that no one was going to participate were included in this stratum.

All people in strata 2 through 7, due to small sample sizes, were eligible to be interviewed in

Waves 2 and 3. In past surveys, a subsample of people were selected for Wave 2 and 3

interviewing.

Every interviewed person in the sportspersons detail sample, after wave 3 interviewing,

received a weight that is the product of the following factors:

1. Screening Weight. This is the person’s final weight from the screening sample.

2. Sportsperson Stratum Adjustment. This factor inflates the weights of persons selected for

the detail sample to account for the subsampling done within each sportsperson

stratum. For 2016, this factor was set to 1 because there was no subsampling.

3. Sportsperson Non-interview Adjustment. This factor adjusts the weights of the

interviewed sportspersons to account for sportspersons selected for the detail sample

for whom no interview was obtained. A person was considered a non-interview if he or

she was not interviewed in the third wave of interviewing.

As shown in Table 6.1, eighteen nonresponse cells were formed within each region/state

defined by the location of the address (MSA or non-MSA), age and sex (3 groups: 16-44

and 45+ for males and 16+ for females) and sportsperson stratum (Fished/Hunted in


P a g e | 53

Wave 1 – stratum 1, Likely to Fish/Hunt – strata 2, 3, and 4, and Unlikely/Very Unlikely to

Fish/Hunt – strata 5, 6, 7).

Table 6.1 Non-interview Cells for Sportspersons

31CBUR Status

Unlikely/Very Fished/Hunted in Likely to Fish/Hunt in

Stratum Unlikely to Fish/Hunt 2016 2016

in 2016

Sex M F M F M F

Age 16-44 45+ 16+ 16-44 45+ 16+ 16-44 45+ 16+

C or B

U or R

4. Sportspersons Ratio Adjustment Factor. This is a ratio adjustment of the detail sample to

the screening sample within the sportspersons sampling strata. This adjustment brings

the population estimates of persons aged 16 years and older from the detail sample into

agreement with the same estimates from the screening sample, which was a much

larger sample. After this adjustment was applied, the final sportsperson weight was

defined.

Wildlife-Watching Sample

Information provided in the screener questionnaire allows stratification of the household

members into strata based on their expected participation in wildlife-watching activities. For

the 2016 FHWAR, the wildlife-watching strata were defined as:

If the person already participated in wildlife-watching activities in 2016, the person was

a wildlife-watching participant and assigned to stratum 0. These people received a detail

interview right after completing the screener questionnaire – also referred to Wave 1

interviewing.

If the person participated in wildlife-watching activities already in 2016 but was not selected for a Wave 1 wildlife-watching interview, the person was a participant and assigned stratum 1.

31Census-defined code assigned to each person based on where the sample housing unit is located – C-Central City of a CBSA, B-Balance of a CBSA, Urbanized Area outside a CBSA, and R-Rural Area outside a CBSA. A CBSA is a Core Based Statistical Area which is a U.S. geographic area defined by the Office of Management and Budget (OMB) that consists of one or more counties (or equivalents) anchored by an urban center of at least 10,000 people plus adjacent counties that are socioeconomically tied to the urban center by commuting.


P a g e | 54

If the person last took trips to participate in wildlife-watching activities in 2015 for 21+ days or spent $300+, the person was defined as an avid participant and was assigned to stratum 2.

If the person last took trips to participate in wildlife-watching activities in 2015 for less than 21 days and spent less than $300, the person was defined as an average participant and assigned to stratum 3.

If the person did not participate in wildlife-watching activities in 2015 but was very likely to in 2016, the person was defined to be an infrequent participant and assigned to stratum 4.

If the person did not participate in wildlife-watching activities in 2015 but was somewhat likely or somewhat unlikely to in 2016, then they were assigned to stratum 5.

If the person was very unlikely to participate in wildlife-watching activities, the person was defined to be a nonparticipant and assigned to stratum 6. Households from the pre-screener that indicated that no one was going to participate were included in this stratum.

Every interviewed person in the wildlife-watching detail sample received a weight that was the

product of the following factors:

1. Screening Weight. This is the person’s final weight from the screening sample.

2. Wildlife-Watching Strata Adjustment. This factor inflates the weights of persons

selected for the detail sample to account for the subsampling done within each wildlife-

watching strata.

3. Wildlife-Watching Non-interview Adjustment. This factor adjusts the weights of the

interviewed wildlife-watching participants to account for wildlife-watching participants

selected for the detail sample for whom no interview was obtained. A person was

considered a non-interview if he or she was not interviewed in the third wave of

interviewing.

As shown in Table 6.2, eight nonresponse cells were formed within each region/state

defined by the location of the address (MSA or non-MSA), age (2 groups: 16-44 and 45+)

and wildlife-watching stratum (Participated before screener interview 1 – strata 0 and

1, Very Likely, Somewhat Likely, or Unlikely/Very Unlikely to Participate – strata 2, 3, 4,

5, and 6).


P a g e | 55

Table 6.2 Non-interview Cells for Wildlife Watching

CBUR Status

Stratum Participated in

2016

Very Likely, Somewhat Likely,

or Unlikely/Very Unlikely to

Participate in 2016

Age 16-44 45+ 16-44 45+

C or B

U or R

4. Wildlife-Watching Ratio Adjustment Factor. This is a ratio adjustment of the detail

sample to the screening sample within the wildlife-watching sampling strata. This

adjustment brings the population estimates of persons aged 16 years and older from

the detail sample into agreement with the same estimates from the screening sample,

which was a much larger sample. After this adjustment was applied, the final wildlife-

watching person weight was defined.

Variance Estimation 2016 FHWAR Design and Methodology Technical Report

P a g e | 56

Chapter 7. VARIANCE ESTIMATION

INTRODUCTION

Since the estimates for the 2016 FHWAR come from a sample, they may differ from an

enumeration of the entire population using the same questionnaires, instructions, and

interviewers. For a given estimator, the difference between an estimate based on a sample and

the estimate that would result if the sample were to include the entire population is known as

sampling error. Variance and standard error (the square root of the variance) are statistical

tools that take into account the magnitude of the sampling error. Although it is imperfect, the

current variance estimation procedure is accurate enough for practical uses of data.

The current approach to estimate the design variances is the successive difference replication

method. The theoretical basis for the successive difference method was discussed by Wolter32

and extended by Fay & Train33 to produce the successive difference replication method, which

has been used widely in most surveys conducted by the Census Bureau.

In the successive difference replication method, the first step in creating a variance estimate is

constructing the replicate factors. The second step is to multiply base weights with replicate

factors to create replicate base weights. The weighting process is then rerun by using each set

of replicate base weights to create final replicate weights. Replicate estimates are created by

using the same estimation method as the original estimate, but applying each set of replicate

weights instead of the original weights. Finally, the replicate and original estimates are used to

compute the variance estimate based on the variability of the replicate estimates and the

original sample estimates. For the FHWAR, the Census Bureau used 160 replicates to calculate

the variance estimates. For additional information on determining the number of replicates,

see http://www.census.gov/prod/2006pubs/tp-66.pdf.34 This chapter describes the

methodology used to produce the direct variance estimates, such that forming 160 replicate

factors and weights and using those weights to compute variance estimates.

Another way to produce variance estimates is to compute generalized variances by using GVFs.

The GVF is a simple model that expresses the variance as a function of the expected value of

the survey estimate. The model’s parameters are estimated by using the direct replicate

variances that are mentioned above. Although the replicate weights have advantages over GVF

models, GVF models are easier to use than replicate weights because these models provide an

easy way to obtain an approximate standard error on numerous characteristics. In addition,

32Wolter, Kirk (1985), Introduction to Variance Estimation, New York: Springer-Verlag New York Inc. 33Fay, Robert, and Train, George (1995), “Aspects of Survey and Model-Based Postcensal Estimation of Income and Poverty Characteristics for States and Counties,” Proceedings of the Section on Government Statistics, American Statistical Association, Alexandria, VA, pp. 154-159. 34Demographic Statistical Methods Division (October 2006), The Current Population Survey: Design and Methodology, U.S. Department of Commerce, U.S. Census Bureau, Technical Paper 66, Washington, D.C.


P a g e | 57

these GVF models have the stability in variance estimation and are more efficient in

computation than using replicate weights. This chapter discusses the GVF models that the

Census Bureau used in the FHWAR survey and estimating generalized standard errors.

REPLICATE FACTORS AND WEIGHTS

The Census Bureau computed replicate weights to estimate variances. In general, the unbiased weights are products of multiplying base weights with special weighting factors, then multiplying these unbiased weights with replicate factors to produce unbiased replicate weights. In addition, the unbiased replicate weights were adjusted by multiplying by non-interview adjustment factors, first-stage adjustment factors, and second-stage ratio adjustment factors to produce the full sample weights. Multiplying these adjustment factors represented the impact of the weighting adjustments on the variance. The replicate factors and weights were created differently for the SR strata and the NSR strata. The Census Bureau derived both sets of replicate weights from methods known as “balanced half-sample” methods. Wolter discussed this methodology and Fay & Train extended the theory. The SR weights were created using the successive difference replication and the NSR weights were created using the modified half sample technique. Replicates for the FHWAR survey were formed through a five-step process:

1. The first step was the construction of a k × k Hadamard matrix, where k is the number of replicates that will be formed.

2. In the second step, each SR case was assigned two rows of the Hadamard matrix and each NSR case one row.

3. In the third step, each sample case used the assigned rows from the Hadamard matrix to calculate its replicate factors.

4. In the fourth step, the replicate factors were multiplied by the full-sample weights to produce the replicate weights.

5. Finally, the full sample and each of the replicate samples went through the weighting process.

Refer to the end of this section for an example to reinforce the steps of the replication method used for FHWAR survey’s replicate weights.

Step 1: Construct the Hadamard Matrix

The first step in creating the replicate weights for the FHWAR survey was the construction of a Hadamard matrix. A Hadamard matrix H is a k × k matrix with all elements either equal to +1 or

-1. Hadamard matrices are unique in that they satisfy k

T

kk k IHH , where I is the identity

matrix of order k, Hk is a k × k Hadamard matrix, and HkT is the transpose of the k × k Hadamard

matrix. The order k is necessarily 1, 2, or 4t, where t is a positive integer. An example of a 2 × 2 Hadamard matrix is as follows:


P a g e | 58

11

112

H (1)

Note that:

H H I2 2 2

T

T

1 1

1 1

1 1

1 1

1 1

1 1

1 1

1 1

2 0

0 22 .

The Hadamard matrix allows the selection of certain replicate samples so that an unbiased estimate of the variance can be calculated with significantly fewer calculations than other half-sample methods. For the FHWAR survey, 160 replicates are used, and thus a 160 × 160 Hadamard matrix is used to form the replicate factors. See Plackett and Burman35 for information on the construction of 160 × 160 Hadamard matrices.

Step 2: Assign Row Values

Assignment of the row values depends on whether the sample case is SR or NSR. As mentioned earlier, replicate weights are formed differently for SR and NSR sample. Each SR case in the full sample will use two rows of the Hadamard matrix and the NSR cases are assigned to one row.

a. Assignment of Row Values for SR Cases Since the first row of most Hadamard matrices consists entirely of +1s, it is not assigned to a sample case. Therefore, the assignment process for the SR cases begins with the assignment of Rows 2 and 3 of the Hadamard matrix to the first sample case. The remaining row assignments are set up to ensure that consecutive sample cases share one row of the Hadamard matrix. Following this algorithm, Rows 3 and 4 are assigned to the second sample case. This row assignment continues until you reach the kth row of the k × k Hadamard matrix. At this point, you skip over the first row and return to the second row for the next assignment. After assigning all the row numbers incrementing by one, continue assigning the row numbers starting from Row 2, but increase the increment interval to two. Using an increment of two, the assignment process will continue with Rows 2 and 4 for the next sample case, followed by Rows 4 and 6, Rows 6 and 8, and so on. Under an increment of two, cycle through the rows twice to pick up all the row numbers. After assigning all increments of two, assign the row numbers with an increment of three. Use three cycles while incrementing by three. Continue to increase the increment and number of cycles up to a maximum increment of ten and then start the assignments over with the increment of one (if the independent sample is large enough to make this necessary). This provides 1,590 unique row assignment pairs.

35 Plackett, R.L. and Burman, J.P. (1946), “The Design of Optimal Multifactorial Experiments,” Biometrika, 33, pp. 305-325.


P a g e | 59

b. Assignment of Row Values for NSR Cases The NSR sampled strata are combined into pseudo-strata within each state to form paired strata. Each pseudo-stratum is assigned to a row of the Hadamard matrix. Within the pseudo-strata, one of the NSR Primary Sampling Units (PSU) is randomly assigned the replicate factor 1.5 and the other NSR PSU receives the factor of 0.5. These values are assigned based on the Hadamard matrix. When the value of the Hadamard matrix changes, the assigned replicate factor changes. For example, if the value of the Hadamard matrix is 1 and the first NSR PSU receives the replicate value of 1.5, the other NSR PSU receives a replicate factor of 0.5. When the value from the Hadamard matrix is –1, the first NSR PSU receives a replicate value of 0.5 and the second NSR PSU receives a replicate value of 1.5. These values are further adjusted to account for the unequal sizes of the original strata within pseudo-stratum. In most cases the pseudo-strata consist of a pair of strata except where an odd number of strata within the state requires that a triplet be formed. In this case two rows of the Hadamard are assigned to the pseudo-stratum resulting in replicate factors of about 0.5, 1.7, and 0.8; or 1.5, 0.3, and 1.2 for the three PSUs assuming roughly equal sizes of the original strata. These values are further adjusted to account for the unequal sizes of the original strata within pseudo-stratum. At the completion of the row assignment, each sample case will have k replicate factors - one factor for each replicate sample.

Step 3: Calculation of the Replicate Factors for the FHWAR Survey

The unique assignment of the row values to the SR sample cases ensures that the replicate factors take on one of three values: 0.3, 1.0, or 1.7. The replicate factors are calculated using the following formula:

.h)2(h)2(1Factor Replicate )r,RII(2

3

)r,RI(2

3

ir ii

(2)

where

i = The sample case (i = 1, 2, …, n) r = The replicate (r = 1, 2, …, k) RI = The first row value assigned to sample case i RII = The second row value assigned to sample case i h(RI, r) = The (RI, r)th cell of the Hadamard matrix h(RII, r) = The (RII, r)th cell of the Hadamard matrix

NOTE: The Hadamard cell to use is determined by the assigned row values and

the column number corresponding to the replicate number. For example,


P a g e | 60

when calculating a replicate factor for replicate 4, use the following cells from the Hadamard matrix: (RI, 4) and (RII, 4).

Step 4: Calculation of the Replicate Weights for the FHWAR Survey

Each case within a probability sample has a sample weight that reflects the inverse of its probability of selection (i.e., the base weight). The weight can be viewed as the number of population members this sample case represents. The fourth step in the replication method calculates the replicate weights for each replicate sample. The replicate weights are calculated using the following formula:

Basewt Factor Replicate = WeightReplicate iirir (3)

where

i = The sample case ( i = 1, 2, …, n), r = The replicate sample (r = 1, 2, …, k),

Replicate Factorir = The replicate factor for the rth replicate of sample case i, and

Basewti = The full-sample base weight of sample case i.

Step 5: Perform the Weighting Process

The final step to create replicate weights for the FHWAR survey involved sending the full sample and each replicate sample through the weighting process. The weighting process involved a series of adjustments to ensure the final estimates were representative of the target population. After the weighting adjustments, the Census Bureau was able to calculate estimates of variance for any FHWAR estimate. The base weights of the FHWAR survey’s sample cases went through the following adjustments:

FHWAR non-interview adjustment.

First-stage ratio adjustment to reduce variances due to the sampling of NSR PSUs.

Second-stage ratio adjustment to reduce variances by controlling to independent estimates of the current population.

Example of the Replication Method

In an attempt to reinforce the steps of the replication method used for the FHWAR survey, the Census Bureau created replicate samples for a sample data set. The sample data set consisted of five cases, all from an SR PSU. Four replicates were created for each sample case. The sample cases and their corresponding full-sample weights were as follows (assume the cases in this example were ordered in a manner reflective of the sample design):


P a g e | 61

Table 7.1 Sample Data for the Replication Method Example

Sample Case Sample Weight

Case #1 15.00

Case #2 23.00

Case #3 19.00

Case #4 16.00

Case #5 21.00

Since four replicates need to be created, a 4 × 4 Hadamard matrix must be constructed. An example of a 4 × 4 Hadamard matrix is as follows:

1+1-1-1+

1-1-1+1+

1-1+1-1+

1+1+1+1+

= H (4)

Recall that each sample case is assigned two rows of the Hadamard matrix. This assignment of rows begins with the second row and allows consecutive sample cases to share a row. The row assignments for the five sample cases are as follows:

Table 7.2 Assignment of Rows for Sample Data

Sample Case Sample Weight Row I Row II

Case #1 30.00 2 3

Case #2 22.00 3 4

Case #3 15.00 4 2

Case #4 20.00 2 4

Case #5 25.00 4 3


P a g e | 62

Applying the Row I and Row II values into formula (2) for case #1, the following replicate factors are calculated:

1.01-22

11-

22

11h2h21RF

1.71-22

11

22

11h2h21RF

.30122

11-

22

11h2h21RF

1.0122

11

22

11h2h21RF

(3,4)2

3

(2,4)2

3

14

(3,3)2

3

(2,3)2

3

13

(3,2)2

3

(2,2)2

3

12

(3,1)2

3

(2,1)2

3

11

Recall that the row value assigned in Table 7.2 and the column number corresponding to the replicate number determines the Hadamard matrix cell to use. Applying the Row I and Row II values in formula (2) for the remaining cases, the replicate factors will be as shown below in Table 7.3.

Table 7.3 Replicate Factors for Sample Data

Replicate Factors

Sample Replicate Replicate Replicate Replicate Sample Case Weight 1 2 3 4

Case #1 30.00 1.0 0.3 1.7 1.0

Case #2 22.00 1.0 1.7 1.0 0.3

Case #3 15.00 1.0 1.0 0.3 1.7

Case #4 20.00 1.0 1.0 1.7 0.3

Case #5 25.00 1.0 0.3 1.0 1.7


P a g e | 63

Now that the replicate factors for each sample case have been calculated, the replicate weights

can be calculated by using formula (3).

Table 7.4 Replicate Weights for Sample Data

Sample Case Sample Weight

Replicate Weights

Replicate 1

Replicate 2

Replicate 3

Replicate 4

Case #1 30.00 30.00 9.00 51.00 30.00

Case #2 22.00 22.00 37.40 22.00 6.60

Case #3 15.00 15.00 15.00 4.50 25.50

Case #4 20.00 20.00 20.00 34.00 6.00

Case #5 25.00 25.00 7.50 25.00 42.50

Total 112.00 112.00 88.90 136.50 110.60

The last step in the creation of the replicate weights is the implementation of any weighting adjustments. In this example, a ratio adjustment is used to control the sample to the population total of 100.00. Therefore, a separate ratio adjustment factor for the full sample and for each replicate sample is calculated. In this example, the ratio adjustment factor formula is as follows:

5

1

00.100

i

i

r

w

RAF (5)

where

i = The sample case (i = 1, 2, …, 5), r = The replicate sample (r = 0, 1, …, 4)

NOTE: Replicate 0 refers to the full sample, wi = The weight for sample case i (either the full-sample weight or a

replicate weight), and RAFr = The ratio adjustment factor for replicate sample r.

Using formula (5), the ratio adjustment factors for the full sample and each replicate sample are calculated as follows:

Full Sample RAF = (100.00 ÷ 112.00) = 0.8929

Replicate 1 RAF = (100.00 ÷ 112.00) = 0.8929

Replicate 2 RAF = (100.00 ÷ 88.90) = 1.1249

Replicate 3 RAF = (100.00 ÷ 136.50) = 0.7326

Replicate 4 RAF = (100.00 ÷ 110.60) = 0.9042


P a g e | 64

To perform the ratio adjustment, multiply the full-sample and replicate weights by the corresponding ratio adjustment factor. The following table provides the ratio adjusted weights.

Table 7.5 Ratio Adjusted Weights for Sample Data

Sample Case

Full Sample Weight

Replicate Weights

Replicate 1

Replicate 2

Replicate 3

Replicate 4

Case #1 26.79 26.79 10.12 37.36 27.12

Case #2 19.64 19.64 42.07 16.11 5.97

Case #3 13.39 13.39 16.87 3.30 23.06

Case #4 17.86 17.86 22.50 24.91 5.42

Case #5 22.32 22.32 8.44 18.32 38.43

Total 100.00 100.00 100.00 100.00 100.00

Using these ratio-adjusted weights, the estimates of variance for survey estimates are ready to be calculated. The next section discusses the calculation of variance estimates using replicates from the replication method.

VARIANCE ESTIMATES

Once the replicate weights of the FHWAR survey were formed, estimates of variance for the full-sample estimate were calculated by using the Fay’s Balanced Repeated Replication (BRR) method36 with the following formula:

) - ( k

4 = )Var(

2

r

k

1r

o oyyy

(6)

where

r = The replicate sample (r = 1......k) o = The full sample k = The total number of replicate samples (k = 160) yo = The survey estimate using the full-sample weights yr = The survey estimate using the replicate weights from replicate r

This variance estimate is the product of a constant and the sum of squared differences between each replicate survey estimate and the full-sample survey estimate. Note that the value of 4 in the equation above arises from the use of successive difference replication. The following example illustrates how a statistic is estimated, replicated, and combined to form a variance estimate. In general, the Census Bureau uses 160 replicate weights to estimate the

36Judkins, D. (1990) “Fay’s Method for Variance Estimation,” Journal of Official Statistics, Vol. 6, No. 3, 1990, pp.223-239.


P a g e | 65

variance, but four replicate weights will be shown in this example to incorporate with the previous example of the replication method.

Example of Replicate Variance Estimation

Recall that the sample consists of five sample cases and four replicates per sample case. The goal of this section is to estimate the total number of hunters in the population and its corresponding estimate of variance. Assume that the five sample cases had the responses shown below in Table 7.6 when asked if they participated in hunting activities during the time of interview.

Table 7.6 Variance Estimation Using Sample Data

Replicate Weights

Hunting Sample Replicate Replicate Replicate Replicate Sample Case Activities Weight 1 2 3 4

Case #1 Yes - 1 26.79 26.79 10.12 37.36 27.12

Case #2 No - 0 19.64 19.64 42.07 16.11 5.97

Case #3 Yes - 1 13.39 13.39 16.87 3.30 23.06

Case #4 Yes - 1 17.86 17.86 22.50 24.91 5.42

Case #5 No - 0 22.32 22.32 8.44 18.32 38.43

To calculate the full-sample survey estimate of the number of people who participate in hunting activities, the full-sample weights of the sample cases that responded “YES” are added to the hunting activities question. Therefore, the total estimate for the number of people who participate in hunting activities is calculated as follows:

Full-Sample of “YES” Responses = 26.79 + 13.39 + 17.86 = 58.04

In order to calculate the variance estimate for this survey estimate, calculate the same survey estimate for each of the replicate samples was required. The replicate survey estimates are as follows:

Replicate 1 of “YES” Responses = 26.79 + 13.39 + 17.86 = 58.04





P a g e | 66

Now use these survey estimates in formula (6) to calculate the variance estimate for the number of people who participate in hunting activities. The calculation of this variance estimate is as follows:

.757.1359536.57009.561025.73

)04.5860.55()04.5857.65()04.5849.49()04.5804.58(4

4

) - ( k

4 = )Var(

2222

2

or

k

1r

o

yyy

Thus Var(yo) = 135.757. Therefore, the survey estimate of “YES” responses to the number of people who participate in hunting activities is 58.04. This survey estimate has an estimated variance of 135.757, or a standard error of 11.65, which is the square root of the estimate of variance.

GENERALIZED VARIANCE FUNCTIONS (GVF)

A GVF is a way to summarize the variances of many different estimates in a simple expression. A GVF assumes that variances for many possible estimates that can be derived from the survey data have the same functional form. The GVF used to estimate the variance of an estimated

population total �̂� is of the form:

𝑉𝑎𝑟(�̂�) = 𝑎�̂�2 + 𝑏�̂� (7)

where a and b are two parameters estimated using least squares regression. The rationale for

this form is the assumption that the variance 𝑉𝑎𝑟(�̂�) can be expressed as the product of the

variance form of a simple random sample for a binomial random variable and a design effect. Denote design effect as deff ; the design effect accounts for the effect of a complex sample design relative to a simple random sample.

Let 𝑃 = 𝑋

𝑁 as the proportion of the population having the characteristic X, where N is the

population size, and let 𝑄 = 1 − 𝑃. The variance of the estimated total �̂�, which is based on a sample of n individuals from the population is:

𝑉𝑎𝑟(�̂�) = 𝑁2 𝑃𝑄(𝑑𝑒𝑓𝑓)

𝑛

= 𝑁2

𝑋

𝑁 (1−

𝑋

𝑁) (𝑑𝑒𝑓𝑓)

𝑛


P a g e | 67

= −(𝑑𝑒𝑓𝑓)𝑋2

𝑛+

(𝑑𝑒𝑓𝑓)𝑁𝑋

𝑛

= −(𝑑𝑒𝑓𝑓) (𝑁

𝑛)

𝑋2

𝑁+

𝑁(𝑑𝑒𝑓𝑓)

𝑛𝑋

Letting 𝑎 = −𝑏

𝑁 and 𝑏 =

𝑁(𝑑𝑒𝑓𝑓)

𝑛

Hence:

𝑉𝑎𝑟(�̂�) = 𝑎�̂�2 + 𝑏�̂�

where N is a control total, so that the variance will equal zero when �̂� = 𝑁. To estimate parameters a and b, the Census Bureau used the model for relative variance (denote Relvar as relative variance), which is shown as below:

𝑅𝑒𝑙𝑣𝑎𝑟 (�̂�) =𝑉𝑎𝑟(�̂�)

�̂�2=

𝑎�̂�2 + 𝑏�̂�

�̂�2= 𝑎 +

𝑏

�̂�

Since 𝑎 = −𝑏

𝑁 the relative variance equation is:

𝑅𝑒𝑙𝑣𝑎𝑟 (�̂�) = 𝑏 (1

�̂�−

1

𝑁)

Using variance estimates obtained directly by using the formula (6), the variance estimates data were fit to the model of relative variance above to estimate parameters a and b by using a regression line. In the FHWAR survey, besides using the equation (7) to estimate hunting and fishing participants in the U.S. population, the following equation was also used to estimate expenditures, days, and trips that relate to hunting and fishing activities:

𝑉𝑎𝑟(�̂�) = 𝑎�̂�2 + 𝑏�̂� +𝑐�̂�2

𝑦 (8)

where �̂� is the size of estimate of expenditures, trips or days, y is the base of the estimate, and a, b and c are the parameters that can be estimated by a relative variance model. After the parameters a, b and c of equations (7) and (8) are determined, the approximate standard error, which is discussed in the next section, can be computed.

ESTIMATING STANDARD ERRORS

The approximate standard error 𝑆�̂� of an estimate �̂� can be obtained with generalized variance parameters a, b and c of equations (7) and (8) and following formulas:


P a g e | 68

𝑆�̂� = √𝑎�̂�2 + 𝑏�̂� (9)

𝑆�̂� = √𝑎�̂�2 + 𝑏�̂� +𝑐�̂�2

𝑦 (10)

Formula (9) is used to calculate the standard errors of levels of hunting, fishing, and wildlife watching. Formula (10) is used for standard errors of aggregates (i.e., trips, days, and expenditures). The sample estimate and its standard error are used to construct a confidence interval. For the FHWAR survey, the Census Bureau constructed 95% confidence intervals for sample

estimates �̂�. The 95% confidence interval indicates that, if the same sampling method was used to select different samples and an interval estimate for each sample was computed, the true population parameter can be expected to fall within the interval estimates 95% of the time.

Example of Estimating Standard Errors and Confidence Intervals

A good example is the previous cycle of the FHWAR; the 2011 survey. Generalized variance parameters are provided in the table below.

Table 7.7 Parameters for Computation of Standard Errors for United States Detail Sportspersons Sample

Characteristic a b c

Sportspersons 16 years and older -0.000070 16,823 -

Days or Trips for Hunters 16 years and older -0.000284 -127,863 46,699

Suppose there were an estimate �̂� = 37,397,000 persons age 16 years and older who either fished or hunted in the U.S in 2011. Using formula (9) and generalized variance parameters in

Table 7.7, the approximate standard error of the estimate �̂� = 37,397,000 sportspersons age 16 years old and older is

S�̂� = √−0.000070 ∗ 37,3970002 + 16,823 ∗ 37,397000 ≈ 728,857

The 95% confidence interval for the estimate �̂� = 37,397,000 sportspersons age 16 years and older is computed as:

37,397,000 ± 1.96 ∗ 728,857 ≈ 35,968,000 𝑡𝑜 38,826,000 Suppose there were an estimate y = 13,674,000 hunters age 16 years and older who engaged in

�̂� = 281,884,000 days of participation in 2011. Using formula (10) with generalized variance parameters in Table 7.7, the approximate standard error of this estimate is:


P a g e | 69

𝑆�̂� = √−0.000284 × 281,884,0002 − 127,863 × 281,884,000 +46,699 × 281,884,0002

13,674,000

≈ 14,586,000

The 95% confidence interval on the estimate �̂� = 281,884,000 days of hunting activities is computed as:

281,884,000 ± 1.96 × 14,586,000 ≈ 253,295,000 𝑡𝑜 310,473,000

In addition, hypothesis tests can be conducted by using standard errors, but hypothesis tests are not covered in the FHWAR survey.

Glossary 2016 FHWAR Design and Methodology Technical Report

P a g e | 70

GLOSSARY AND DESCRIPTIONS OF ACRONYMS

AAPOR American Association for Public Opinion Research ACS American Community Survey

AFWA Association of Fish and Wildlife Agencies CAPI Computer Assisted Personal Interview CATI Computer Assisted Telephone Interview

CARRA Center for Administrative Records Research and Applications CAUS Community Address Updating System CBUR Census-defined code assigned to each person based on where the sample housing

unit is located – C-Central City of a CBSA, B-Balance of a CBSA, Urbanized Area Outside a CBSA, and R-Rural Area outside a CBSA.

CBSA Core Based Statistical Area. A U.S. geographic area defined by the Office of Management and Budget (OMB) that consists of one or more counties (or equivalents) anchored by an urban center of at least 10,000 people plus adjacent counties that are socioeconomically tied to the urban center by commuting.

CPS Current Population Survey CV Coefficient of Variation

DAAL Demographic Area Address Listings DEFF Design Effects

DSF Delivery Sequence File DSMD Demographic Statistical Methods Division

EDS Exclude from Delivery Statistics FHWAR National Survey of Fishing, Hunting, and Wildlife-Associated Recreation

FIPS Federal Information Processing Standards FR Field Representative. The Census Bureau staff assigned to one of the six Census

Bureau regional offices. FRs conduct interviews primarily in person (personal visit) utilizing either paper questionnaires or a CAPI instrument.

FWS U.S. Department of the Interior, Fish and Wildlife Service GEO Geography Division GIS Geographic Information System

GPS Geographic Positioning System GQ Group Quarters

GVF Generalized Variance Function HU Housing Unit IAA Interagency Agreement IDS Include in Delivery Statistics

LACS Locatable Address Conversion System LUCA Local Update of Census Addresses MAF Master Address File

MAFID Master Address File Identifier

MSA Metropolitan Statistical Area

MTdb MAF/TIGER Database

Glossary 2016 FHWAR Design and Methodology Technical Report

P a g e | 71

NPC National Processing Center – Division of the Census Bureau that processes letter mailouts, labeling operations, and editing and keying of paper questionnaires. Located in Jeffersonville, Indiana.

NSR Non Self-Representing PSU Primary Sampling Unit SAS Software used to edit, clean, estimate, and analyze the data

SI Sampling Interval SR Self-Representing

TIGER Topologically Integrated Geographic Encoding and Referencing System UAA Undeliverable As Addressed

UFUF Unit Frame Universe Files USPS United States Postal Service VHU Valid Housing Units

Appendix A – 2016 FHWAR Materials 2016 FHWAR Design and Methodology Technical Report

P a g e | 72

APPENDIX A. 2016 FHWAR MATERIALS

ADVANCE LETTERS

Advance letters were mailed to respondents to invite them to participate in the survey.


P a g e | 73


P a g e | 74


P a g e | 75


P a g e | 76


P a g e | 77


P a g e | 78

FREQUENTLY ASKED QUESTIONS

The Frequently Asked Questions were included with all advance letter mailings.


P a g e | 79

QUICK FACTS

The Quick Facts brochure was included with the advance letter mailings for the pre-screener,

Wave 1, and Wave 2.


P a g e | 80

REFERENCE AID

The Reference Aid was included with the advance letter mailings for Waves 1, 2, and 3. The

Reference Aid was a collection of response options from a variety of questions in the

questionnaire. The Aid was used to inform the household of the types of questions they would

be asked in the upcoming interview and to help remind them of activity they participated in or

purchases they may have made since January 1, 2016.


P a g e | 81


P a g e | 82


P a g e | 83


P a g e | 84


P a g e | 85


P a g e | 86


P a g e | 87

Appendix B – 2016 Questionnaires 2016 FHWAR Design and Methodology Technical Report and Survey Questions

P a g e | 88

APPENDIX B. 2016 QUESTIONNAIRES AND SURVEY QUESTIONS

PRE-SCREENER QUESTIONNAIRE


P a g e | 89


P a g e | 90


P a g e | 91


P a g e | 92

ABDRIGED LIST OF WAVES 1 AND 2 SCREENER INTERVIEW QUESTIONS

SCINTRO First, I will ask you about the people in your household who may or may not participate in wildlife-associated activities. List all persons staying there and all persons who usually live there who are absent. Start with the name of the person (or one of the persons) who owns or rents the residence.

SCAGE What is ARE_IS age?

SCSEX What is YOUR_NAMES sex?

SCRELATE What is YOUR_NAMES relationship to YOU_REFNAME?

SCMSTAT ARE_IS_CAP now - married, widowed, divorced, separated OR never

married?

SCSCHOOL What is the highest grade (or year) of regular school you have/Name has ever completed?

SCJOB DO_DOES have a job or business?

SCRETIRE ARE_IS_CAP retired, going to school, keeping house or doing something

else?

SCHISP ARE_IS_ANY of Spanish, Hispanic, OR Latino origin?

SCRACE What is YOUR_NAMES race?

SCEVHUNT Now I would like to ask you about YOUR_HUNTING ACTIVITIES_MEMBERS. HAVEYOU_HASANYONE EVER hunted game or other wildlife?

SCHUNT96 HAVE_HAS YOU_96 done any hunting so far in 2016?

SCHUNT95 Did YOU_95 hunt game or other wildlife IN2015?

SCHU95FY Was 2015 the first year that YOU_95FY hunted?

SCH95DAY During 2015, did YOU_95DAY hunt 10 or more days?


P a g e | 93

SCHUSPND During 2015, did YOU_SCHUSPND spend $100 or more for hunting, that is, YOUR_HIS_HER share of expenses (for equipment, travel, lodging, license fees, etc.)?

SCHNTREC What was the most recent yearBEFORE_HUNT_NOT_2016 in which

YOU_NAME hunted?

SCH96LIK On a scale of 1 to 5, where 1 is "Very Unlikely" and 5 is "Very Likely," how likely is it that YOU_NAME will do any hunting during 2016?

TARGET Did YOU_TARGET participate in target shooting or sport shooting with a

firearm (i.e., - rifle, shotgun, muzzleloader, handgun, air gun) in 2015, not including hunting? Please include any informal target shooting or sport shooting.

ARCHERY Did YOU_ARCHERY participate in archery activities using a bow and arrow,

compound bow, or crossbow in 2015, not including hunting? Please include any informal archery activity.

SCEVFISH Now I would like to ask you about YOUR_FISHING

ACTIVITIES_MEMBERS. HAVEYOU_HASANYONE EVER done any recreational fishing, including shellfishing?

SCFISH96 HAVE_HAS YOU_96 done any fishing so far in 2016?

SCFI95FY Was 2015 the first year that YOU_95FY fished?

SCF95DAY During 2015, did YOU_95DAY fish 10 or more days?

SCFISPND During 2015, did YOU_SCFISPND spend $100 or more for fishing, that is,

YOUR_HIS_HER share of expenses (for equipment, travel, lodging, license fees, etc.)?

SCFSHREC What was the most recent yearBEFORE_FISH_NOT_2016 in which

YOU_NAME fished?

SCF96LIK On a scale of 1 to 5, where 1 is "Very Unlikely" and 5 is "Very Likely," how likely is it that YOU_NAME will do any fishing during 2016?


P a g e | 94

SCWILD96 Next, I would like to ask about SPECIAL INTEREST in wildlife in ways OTHER THAN hunting and fishing. We are interested in whether you closely observe, photograph, feed, or maintain natural areas or plantings for wildlife. Please do not include noticing wildlife while doing other activities. Do not include trips to zoos, circuses, aquariums, museums, or scouting for game. By wildlife I mean birds, mammals, fish, insects, reptiles such as snakes and lizards, and amphibians such as frogs. DO NOT include farm animals and pets. HAVE_HAS YOU_96 taken any SPECIAL INTEREST in wildlife so far in 2016?

SCWILD95 Did YOU_95 take special interest in wildlife IN2015? Do not include trips to

zoos, circuses, aquariums, museums, or scouting for game.

SCW95DAY During 2015, did YOU_95DAY observe, photograph, or feed wildlife at least one mile from home for 21 or more days?

SCWWSPND During 2015, did YOU_SCWWSPND spend $300 or more to observe,

photograph, or feed wildlife at least one mile from home, that is, YOUR_HIS_HER share of expenses (for equipment, travel, lodging, license fees, etc.)

SCW96LIK On a scale of 1 to 5, where 1 is "Very Unlikely" and 5 is "Very Likely," how

likely is it that YOU_NAME will take SPECIAL INTEREST in wildlife this year?

SCINC SCINC_FILL What was the total income of this household during 2015 before taxes and other deductions?


P a g e | 95

ABDRIGED LIST OF WAVE 3 DETAIL INTERVIEW QUESTIONS

SINTRO I would like to ask you some questions about hunting and fishing. These questions will deal with the kinds of hunting or fishing you did, where you went, and what you bought. All questions refer to the time period between Date of last Sportsperson interview and December 31, 2016.

RECFISH Did you do any recreational fishing, including shellfishing, in the United

States from Date of last Sportsperson interview to December 31, 2016? Please do not include as fishing occasions when you only observed others fish.

USHUNT Did you hunt in the United States from Date of last Sportsperson interview

to December 31, 2016? Please do not include as hunting occasions when you only observed others hunt or when you only scouted.

HUNT_ST In which state or states did you hunt?

USDAYS_H From Date of last Sportsperson interview to December 31, 2016, how many

days did you hunt in the United States?

STDAYS_H How many days did you hunt in Hunt state?

HUNT Please tell me which kinds of game you hunted over this period in Hunt state.

BG_TRIP You reported hunting Game fill for big game in Hunt state which we consider to be big game. How many trips lasting a single day or multiple days did you take from Date of last Sportsperson interview to December 31, 2016 ^HUNT_IN_FLAG Hunt state to hunt big game?

BG_DAYS How many days in Hunt state did you hunt big game?

BGDIFDAY How many days did you hunt Big Game species day fill in Hunt state?

BG_PRVT Did you do any big game hunting in Hunt state on privately owned land?

BGDYPRV How many days?

BGPUBLIC Did you do any big game hunting in Hunt state on land owned by the local,


P a g e | 96

State, or Federal Government?

BGDYSPUB How many days?

SM_TRIP You reported hunting Game fills for small game in Hunt state which we consider to be small game. How many trips lasting a single day or multiple days did you take from Date of last Sportsperson interview to December 31, 2016 in/to Hunt state to hunt small game?

SM_DAYS How many days in Hunt state did you hunt small game?

SMDIFDAY How many days did you hunt Show small game species in Hunt state? Fill

small game text

SM_PRVT Did you do any small game hunting in Hunt state on privately owned land?

SMDYPRV How many days?

SMPUBLIC Did you do any small game hunting in Hunt state on land owned by the local, State, or Federal Government?

SMDYSPUB How many days?

MB_TRIP You reported hunting

Fill of migratory bird species in Hunt state which we consider to be migratory birds. How many trips lasting a single day or multiple days did you take from Date of last Sportsperson interview to December 31, 2016 in/to Hunt state to hunt migratory birds?

MB_DAYS How many days in Hunt state did you hunt migratory birds?

MB_COMBO How many days did you hunt waterfowl (geese and/or ducks) in Hunt state?

MBDIFDAY How many days did you hunt List of migratory game species in Hunt state?

MB_PRVT Did you do any migratory bird hunting in Hunt state on privately owned

land?

MBDYPRV How many days?

MBPUBLIC Did you do any migratory bird hunting in Hunt state on land owned by the


P a g e | 97

local, State, or Federal Government?

MBDYSPUB

How many days?

OA_TRIP You reported hunting Game fill for other animals in Hunt state which we consider to be other animals. How many trips lasting a single day or multiple days did you take from Date of last Sportsperson interview to December 31, 2016 in/to Hunt state to hunt these other animals?

OA_DAYS

How many days in Hunt state did you hunt these other animals?

OA_PRVT

Did you hunt these other animals in Hunt state on privately owned land?

OADYPRV

How many days?

OAPUBLIC Did you hunt these other animals in Hunt state on land owned by the local, State, or Federal Government?

OADYSPUB

How many days?

INTRO1 Now I would like to ask you some questions about your hunting-related expenditures from Date of last Sportsperson interview to December 31, 2016. If you paid for others or if someone else paid for you, please include ONLY YOUR SHARE of the cost. Do not include amounts paid for license fees, stamps, tags, or equipment purchases. I will ask about those costs later.

The following fields (BGSHAR_A through BGSHAR5_1 were asked for each Game, Small Game, Migratory Birds, and Other Animals).

type of hunting (Big

BGSHAR_A When you were hunting in Hunt state CHIEFLY for BIG GAME from Date of

last Sportsperson interview to December 31, 2016, how much was spent for YOUR SHARE of - Food, drink, and refreshments?

BGSHAR_B

Lodging at motels, cabins, lodges, or campgrounds, etc.?

BGSHAR_C

Public transportation by airplane?

BGSHAR_D Other public transportation, including trains, buses, and car rentals, etc.?


P a g e | 98

BGSHAR_E

BGSHAR2_F

BGSHAR2_G

BGSHAR2_H

BGSHAR2_I

BGSHAR2_J

BGSHAR3

BGSHAR4_K

BGSHAR4_L

BGSHAR4_M

BGSHAR5_1

The round-trip cost for transportation by private vehicle? Do not include boating expenses. When you were hunting in Hunt state CHIEFLY for BIG GAME from Date of last Sportsperson interview to December 31, 2016, how much was spent for YOUR SHARE of - Guide fees, pack trip or package fees? Public land use or access fees? Do NOT include leases. Private land use or access fees? Do NOT include leases. Heating and cooking fuel? Equipment rental such as boats, hunting or camping equipment, etc.? Did you have ANY boating expenses boat rental fill while big game hunting in Hunt state? How much for... Boat Fuel? Boat launching fees? Do not include land access fees already reported. Boat mooring, storage, maintenance, pumpout fees, and insurance? The total amount you spent on your big game hunting trip(s) to Hunt state was Total big game expenses, not including airfare. How much of this was spent in your resident state of Resident state?

The following hunting.

questions were asked once the trip expenditures were asked for each type of

MUZZHNT

FIREHUNT

The next series of questions will refer to ALL of your hunting for any species from Date of last Sportsperson interview to December 31, 2016. Did you do any hunting with a muzzleloader? Any hunting with a firearm other than a muzzleloader, such as a shotgun or rifle?


P a g e | 99

HUNTBOW Any hunting with a bow and arrow?

DAYMUZZ How many days did you hunt with a muzzleloader from Date of last Sportsperson interview to December 31, 2016?

DAYFIRE How many days did you hunt with a firearm other than a muzzleloader

from Date of last Sportsperson interview to December 31, 2016?

DAYBOW How many days did you hunt with a bow and arrow from Date of last Sportsperson interview to December 31, 2016?

FISH_ST Now I would like you to think about all the fishing you did

from Date of last Sportsperson interview to December 31, 2016. This includes Great Lakes, other freshwater, and saltwater fishing. In which state or states did you fish?

USDAYS_F From Date of last Sportsperson interview to December 31, 2016, how many

days did you fish in the United States?

STDAYS_F From Date of last Sportsperson interview to December 31, 2016 how many days did you fish in Fishing state?

SSTATE Did you do any recreational saltwater fishing in Fishing state?

GSTATE Did you do any Great Lakes fishing in Fishing state?

FSTATE Did you do any freshwater fishing in Fishing state?

DAYFRESH From Date of last Sportsperson interview to December 31, 2016, how many

days did you fish in any freshwater, including the Great Lakes, in New York?

WHCHGL In which of the Great Lakes, including their tributaries and connecting waters, did you fish?

DAYS_GL How many days did you fish in Great Lakes names?

GL_TRIP How many trips lasting a single day or multiple days did you take in/to

Fishing state to go Great Lakes fishing?

GLSTDAYS From Date of last Sportsperson interview to December 31, 2016, how many days in Fishing state did you go Great Lakes fishing?


P a g e | 100

GLTYP On your Great Lakes trip(s) over this period in/to Fishing state, what types

of fish were you primarily fishing for? Please do not report what you caught unintentionally.

GLDAYS How many days did you fish for Great Lakes species name fill in Fill text in

GLDAYS?

BOATGL Did you fish from a boat while Great Lakes fishing in Fishing state?

BOATGLDY How many days did you Great Lakes fish from a boat in Fishing state?

FR_TRIP How many trips lasting a single day or multiple days did you take in/to Fishing state to go freshwater fishing?

FR_DAYS How many days in Fishing state did you go freshwater fishing?

FRTYP On your trip(s) over this period in/to Fishing state, what types of fish were

you primarily fishing for? Please do not report what you caught unintentionally.

DAYS_FR How many days did you fish for Freshwater species name for DAYS_FR

FR_POND Did you fish in Fishing state in ponds or lakes or reservoirs?

DAYPOND How many days?

FR_RIVER Did you fish in Fishing state in rivers or streams?

RIVERDAY How many days?

BOATFR Did you fish from a boat while FRESHFILL fishing in Fishing state?

BOATFRDY How many days did you FRESHFILL fish from a boat in Fishing state?

SALTTRIP How many trips lasting a single day or multiple days did you take in/to

Fishing state to go saltwater fishing?

SALTDAYS How many days in Fishing state did you go saltwater fishing?

CRABS1 On the day you went saltwater fishing in Fishing state, were you seeking ONLY crabs, clams, or other shellfish, but NOT finfish?


P a g e | 101

FINFISH1 On the day you went saltwater fishing in Fishing state, were you finfishing ONLY?

CRABS2 How many of your Fill SALTDAYS if non DK/RF saltwater fishing days in

Fishing state were you seeking ONLY crabs, clams, or other shellfish, but NOT finfish?

FINFISH2

Of your Fill SALTDAYS if non DK/RF saltwater fishing days in Fishing state, how many were for finfishing ONLY?

SLTYP

On your trip(s) from Date of last Sportsperson interview to December 31, 2016 in/to Fishing state, what types of fish were you primarily fishing for? Please do not report what you caught unintentionally.

SALTDAY

How many days did you fish for Saltwater species name fill in Fishing state?

BOATSL

Did you fish from a boat while saltwater fishing in Fishing state?

BOATSLDY

How many days did you saltwater fish from a boat in Fishing state?

INTRO2 Now I would like to ask you some questions about your fishing-related expenditures from Date of last Sportsperson interview to December 31, 2016. If you paid for others, or if someone else paid for you, please include ONLY YOUR SHARE of the cost. Do not include amounts paid for license fees, stamps, tags, or equipment purchases. I will ask about those costs later.

The following fields (GLSHAR_A through GLSHAR5_1 were asked for each Lakes, Other Freshwater, and Saltwater).

type of fishing (Great

GLSHAR_A When you were fishing in Fishing state CHIEFLY in the Great Lakes from

Date of last Sportsperson interview to December 31, 2016, how much was spent for YOUR SHARE of - Food, drink, and refreshments?

GLSHAR_B

Lodging at motels, cabins, lodges, or campgrounds, etc.?

GLSHAR_C

Public transportation by airplane?

GLSHAR_D

Other public transportation, including trains, buses, and car rentals, etc.?

GLSHAR_E The round-trip cost for transportation by private vehicle?


P a g e | 102

GLSHAR_F

GLSHAR2_G

GLSHAR2_H

GLSHAR2_I

GLSHAR2_J

GLSHAR2_K

GLSHAR2_L

GLSHAR3

GLSHAR4_M

GLSHAR4_N

GLSHAR4_O

GLSHAR5_1

Do not include boating expenses. Guide fees, pack trip or package fees (incl. fees for party and charter boats, etc.)? When you were fishing in Fishing state CHIEFLY in the Great Lakes from Date of last Sportsperson interview to December 31, 2016, how much was spent for YOUR SHARE of - Public land use or access fees? Do NOT include leases. Private land use or access fees? Do NOT include leases. Bait (live, cut, prepared)? Do NOT include lures. Ice? Heating and cooking fuel? Equipment rental such as boats, fishing or camping equipment, etc.? Did you have ANY boating expenses boat rental fill during these trips in/to Fishing state? How much for... Boat Fuel? Boat launching fees? Do not include land access fees already reported. Boat mooring, storage, maintenance, pumpout fees, and insurance? The total amount you spent on your Great Lakes fishing trip(s) to Fishing state was Cost of Great Lake expenditures, not including airfare. How much of this was spent in your resident state of Resident state?

The following fishing.

questions were asked once the expenditures above were asked for each type of

FLYFISH

Now I would like to ask you a few questions about the fishing methods you may have used in the United States. From Date of last Sportsperson interview to December 31, 2016, did you


P a g e | 103

flyfish?

DAYSFLY How many days?

ICEFISH From Date of last Sportsperson interview to December 31, 2016, did you icefish?

DAYS_IF How many days?

H_EQP_A As I read the following list, tell me those items that you bought for yourself

or that were bought for you PRIMARILY FOR HUNTING. From Date of last Sportsperson interview to December 31, 2016, did you purchase or acquire - Rifles?

H_EQP_B Shotguns?

H_EQP_C Muzzleloaders or other so-called primitive firearms?

H_EQP_D Pistols, handguns?

H_EQP_E Bows, arrows, or other archery equipment?

H_EQP_F Telescopic sights?

H_EQP2_G From Date of last Sportsperson interview to December 31, 2016, did you

purchase or acquire - Decoys, game calls?

H_EQP2_H Ammunition?

H_EQP2_I Hand loading equipment and components (i.e., powder, shot, etc.)?

H_EQP2_J Hunting dogs and associated costs?

H_EQP2_K Any other purchases (such as cases and carriers for equipment or game,

hunting knives, etc.)?

H_EQP2_K1 What was that purchase?


P a g e | 104

RIFLECST What was the total cost of the RIFLE(S) purchased?

RIFLEUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

RIFLE_ST In which state(s) did you purchase this equipment?

GUNSCOST What was the total cost of the SHOTGUN(S) purchased?

GUNS_USE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

GUNS_ST In which state(s) did you purchase this equipment?

MUZZCST What was the total cost of the MUZZLELOADER(S) OR OTHER SO-CALLED PRIMITIVE FIREARM(S) purchased?

MUZZUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

MUZZ_ST In which state(s) did you purchase this equipment?

PISTLCST What was the total cost of the PISTOLS, HANDGUNS purchased?

PISTLUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

PISTL_ST In which state(s) did you purchase this equipment?

BOWSCOST What was the total cost of the BOWS, ARROWS, OR OTHER ARCHERY EQUIPMENT purchased?

BOWS_USE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

BOWS_ST In which state(s) did you purchase this equipment?

SCOPECST What was the total cost of the TELESCOPIC SIGHT(S) purchased?

SCOPEUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

SCOPE_ST In which state(s) did you purchase this equipment?


P a g e | 105

DECOYCST What was the total cost of the DECOYS, GAME CALLS purchased?

DECOYUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory

birds, OR other animals?

DECOY_ST In which state(s) did you purchase this equipment?

AMMCOST What was the total cost of the AMMUNITION purchased?

AMM_USE Is it PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

AMMO_ST In which state(s) did you purchase the AMMUNITION?

HANDCST What was the total cost of the HAND LOADING EQUIPMENT AND

COMPONENTS purchased?

HANDUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

HAND_ST In which state(s) did you purchase this equipment?

DOGSCOST What was the total cost of the HUNTING DOGS AND ASSOCIATED COSTS?

DOGSUSE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory


DOGS_ST In which state(s) did you make this purchase?

OTH_COST What was the total cost of the Entry in H_EQP2_K1 purchased?

OTH_USE Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

HOTHST In which state(s) did you purchase this equipment?

F_EQP_A As I read the following list, tell me those items that you bought for yourself

or that were bought for you PRIMARILY FOR FISHING. From Date of last Sportsperson interview to December 31, 2016, did you purchase or acquire -


P a g e | 106

Rods, reels, poles, and rod making components?

F_EQP_B Lines and leaders?

F_EQP_C Artificial lures, flies, baits, and dressing for flies or lines?

F_EQP_D Hooks, sinkers, swivels, and other items attached to a line, except lures and baits?

F_EQP2_E From Date of last Sportsperson interview to December 31, 2016, did you

purchase or acquire - Tackle boxes?

F_EQP2_F Creels, stringers, fish bags, landing nets and gaff hooks?

F_EQP2_G Minnow traps, seines, and bait containers?

F_EQP2_H Depth finders, fish finders, and other electronic fishing devices?

F_EQP2_I Ice fishing equipment (such as tip-ups and tilts, ice fishing houses, etc.)?

F_EQP2_J Any other purchases (such as scales, knives, fishing hook disgorgers, fish

fighting chairs, outriggers, downriggers, rod holders and rod belts, fishing vests, and spear fishing and scuba equipment)?

F_EQP2_J1 What was that purchase?

RODSCOST What was the total cost of the RODS, REELS, POLES, AND ROD MAKING

COMPONENTS purchased?

RODSUSE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

RODSST In which state(s) did you purchase this equipment?

LINECOST What was the total cost of the LINES AND LEADERS purchased?

LINE_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

LINEST In which state(s) did you purchase this equipment?


P a g e | 107

LURECOST What was the total cost of the ARTIFICIAL LURES, FLIES, BAITS, OR DRESSING FOR FLIES AND LINES purchased?

LURE_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other

freshwater, OR saltwater?

LUREST In which state(s) did you purchase this equipment?

HOOKCOST What was the total cost of the HOOKS, SINKERS, SWIVELS, AND OTHER ITEMS ATTACHED TO A LINE, EXCEPT LURES AND BAITS purchased?

HOOK_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


HOOKST In which state(s) did you purchase this equipment?

TACKCOST What was the total cost of the TACKLE BOX(ES) purchased?

TACK_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

TACKST In which state(s) did you purchase this equipment?

CREELCST What was the total cost of the CREELS, STRINGERS, FISH BAGS, LANDING

NETS, AND GAFF HOOKS purchased?

CREELUSE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

CRL_ST In which state(s) did you purchase this equipment?

BAITCOST What was the total cost of the MINNOW TRAPS, SEINES, AND BAIT

CONTAINERS purchased?

BAIT_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

BAITST In which state(s) did you purchase this equipment?

FINDCOST What was the total cost of the DEPTH FINDERS, FISH FINDERS, AND OTHER

ELECTRONIC FISHING DEVICES purchased?

FIND_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


P a g e | 108


FINDST In which state(s) did you purchase this equipment?

ICE_COST What was the total cost of the ICE FISHING EQUIPMENT (SUCH AS TIP-UPS AND TILTS, ICE FISHING HOUSES, ETC.) purchased?

ICE_USE Is it/Are they PRIMARILY for use in fishing in the Great Lakes or other

freshwater?

ICE_ST In which state(s) did you purchase this equipment?

OTHERCST What was the total cost of the Entry in F_EQP2_J1 purchased?

OTHERUSE Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

FOTHST In which state(s) did you purchase this equipment?

FH_EQP_A As I read the following list, tell me those items that you bought for yourself

or that were bought for you PRIMARILY FOR USE IN EITHER FISHING OR HUNTING. From Date of last Sportsperson interview to December 31, 2016, did you purchase or acquire - Camping equipment (such as sleeping bags, packs, duffel bags, tents, etc.)?

FH_EQP_B Binoculars, field glasses, telescopes, etc.?

FH_EQP_C Special fishing or hunting clothing, foul weather gear, boots, waders, etc.?

FH_EQP_D Processing and taxidermy costs?

FH_EQP2_E From Date of last Sportsperson interview to December 31, 2016, did you

purchase or acquire - Books, magazines, or DVDs devoted to fishing or hunting?

FH_EQP2_F Dues or contributions to national, state, or local organizations?

FH_EQP2_G Any other purchases (such as GPS devices, snowshoes, skis, maintenance

and repair of equipment, etc.)? DO NOT INCLUDE BOATS OR VEHICLES.


P a g e | 109

FH_EQP2_G1 What was that purchase?

CAMPCOST What was the total cost of the CAMPING EQUIPMENT purchased?

CAMP_USE Is it PRIMARILY for use in fishing OR for use in hunting?

CAMPFISH Is it PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR

saltwater?

CAMPHNT Is it PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

CAMPST In which state(s) did you purchase this equipment?

BINOCOST What was the total cost of the BINOCULARS, FIELD GLASSES, TELESCOPES,

ETC. purchased?

BINO_USE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

BINOFISH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

BINOHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory


BINOST In which state(s) did you purchase this equipment?

GEARCOST What was the total cost of the SPECIAL FISHING OR HUNTING CLOTHING, FOUL WEATHER GEAR, BOOTS, WADERS, ETC. purchased?

GEAR_USE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

GEARFISH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


GEARHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

GEARST In which state(s) did you purchase this equipment?

TAXICOST What were the total PROCESSING AND TAXIDERMY COSTS?


P a g e | 110

TAXI_USE Were the costs PRIMARILY for fishing OR hunting?

TAXIFISH Were the costs PRIMARILY for fishing in the Great Lakes, other freshwater, OR saltwater?

TAXIHNT Were the costs PRIMARILY for hunting big game, small game, migratory


TAXIST In which state(s) did you pay these costs?

BOOKCOST What was the total cost of the BOOKS, MAGAZINES, AND DVD'S DEVOTED TO FISHING OR HUNTING purchased?

BOOK_USE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

BOOKFISH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


BOOKHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

BOOKST In which state(s) did you make this purchase?

DUECOST What was the total cost of the DUES OR CONTRIBUTIONS TO NATIONAL,

STATE, OR LOCAL ORGANIZATIONS?

DUES_USE Are they PRIMARILY for use in fishing OR for use in hunting?

DUESFISH Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

DUESHNT Are they PRIMARILY for use in hunting big game, small game, migratory


DUESST In which state(s) did you pay these costs?

FHCOST What was the total cost of the Entry in FH_EQP2_G1 purchased?

FH_USE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

FH_FISH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?


P a g e | 111

FH_HNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

FH_ST In which state(s) did you purchase this equipment?

EqpAnn_A As I read the following list, tell me those items that you bought for yourself

or that were bought for you PRIMARILY FOR USE IN EITHER FISHING OR HUNTING. During 2016, did you purchase or acquire - A bass boat?

EqpAnn_B Another type of motor boat?

EqpAnn_C A canoe or other non-motor boat?

EqpAn2_D A boat motor, boat trailer or hitch, or other boat accessories?

EqpAn2_E A pickup, camper, van, travel or tent trailer, motor home, house trailer,

recreational vehicle (RV)?

EqpAn2_F A cabin?

EqpAn2_G A trail bike, dune buggy, 4 x 4 vehicle, ATV, 4-wheeler, snowmobile?

EqpAn2_H Anything else, including airplanes, freezers, etc.?

AnnOth_H What was that purchase?

BASSCOST What was the amount paid for the BASS BOAT in 2016? Please only report the cost in 2016, not the total value of the item.

BASSUSE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

BASSFISH Is it PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR

saltwater?

BASSHNT Is it PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

BASST In which state(s) did you purchase this equipment?


P a g e | 112

BOATCOST What was the amount paid for the OTHER TYPE OF MOTOR BOAT in 2016? Please only report the cost in 2016, not the total value of the item.

BOATUSE Is it PRIMARILY for use in fishing OR for use in hunting?

BOATFISH Is it PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR

saltwater?

BOATHNT Is it PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

BOATST In which state(s) did you purchase this equipment?

CANOECST What was the amount paid for the CANOE OR OTHER NON-MOTOR BOAT in

2016? Please only report the cost in 2016, not the total value of the item.

CANOEUSE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

CANOEFSH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

CANOEHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory


CANOEST In which state(s) did you purchase this equipment?

MOTORCST What was the amount paid for the BOAT MOTOR, BOAT TRAILER/HITCH OR OTHER BOAT ACCESSORIES in 2016? Please only report the cost in 2016, not the total value of the item.

MOTORUSE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

MOTORFSH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


MOTORHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

MOTORST In which state(s) did you purchase this equipment?

VANCOST What was the amount paid for the PICKUP, CAMPER, VAN, TRAVEL OR TENT

TRAILER, MOTOR HOME, HOUSE TRAILER, RECREATIONAL VEHICLE in 2016? Please only report the cost in 2016, not the total value of the item.


P a g e | 113

VANUSE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

VANFISH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


VANHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

VANST In which state(s) did you purchase this equipment?

CABINCST What was the amount paid for the CABIN in 2016?

Please only report the cost in 2016, not the total value of the item.

CABINUSE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

CABINFSH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

CABINHNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory


CABINST In which state(s) did you purchase this equipment?

OFF_COST What was the amount paid for the TRAIL BIKE, DUNE BUGGY, 4 X 4 VEHICLE, ATV, 4-WHEELER, OR SNOWMOBILE in 2016? Please only report the cost in 2016, not the total value of the item.

OFFUSE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?

OFF_FISH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other


OFF_HNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory birds, OR other animals?

OFF_ST In which state(s) did you purchase this equipment?

EQPCST What was the amount paid for the Entry in ANNOTH_H in 2016?


OTHG_USE Is it/Are they PRIMARILY for use in fishing OR for use in hunting?


P a g e | 114

OTHG_FSH Is it/Are they PRIMARILY for use in fishing in the Great Lakes, other freshwater, OR saltwater?

OTHG_HNT Is it/Are they PRIMARILY for use in hunting big game, small game, migratory


OTHG_ST In which state(s) did you purchase this equipment?

HOWN In 2016, did you own land in the United States, in part or whole, PRIMARILY for the purpose of hunting?

HOWN_ACRE How many acres did you own?

HOWN_CLUB Were you part of a club or group, including a family group, owning this land

in 2016?

HOWN_CNUM How many others were members of this group or club, not counting yourself?

HOWN_SHR What did you spend in 2016 for YOUR SHARE of the land which was owned

PRIMARILY for hunting? Include mortgage, taxes, maintenance, and down payment cost if purchased in 2016. Do not include the cost of a cabin if reported earlier.

HOWN_ST In which state(s) was this land located?

HLEASE In 2016, did you lease land in the United States, alone or with others,

PRIMARILY for the purpose of hunting?

HLSE_ACRE How many acres did you lease?

HLSE_CLUB Were you part of a club or group, including a family group, leasing this land in 2016?

HLSE_CNUM How many others were members of this group or club, not counting

yourself?

HLSE_SHR What did you spend in 2016 for YOUR SHARE of the land which was leased PRIMARILY for hunting?

HLSE_ST In which state(s) was this land located?

FOWN In 2016, did you own land in the United States, in part or whole, PRIMARILY


P a g e | 115

for the purpose of fishing?

FOWN_ACRE How many acres did you own?

FOWN_CLUB Were you part of a club or group, including a family group, owning this land in 2016?

FOWN_CNUM How many others were members of this group or club, not counting

yourself?

FOWN_SHR What did you spend in 2016 for YOUR SHARE of the land which was owned PRIMARILY for fishing? Include mortgage, taxes, maintenance, and down payment cost if purchased in 2016. Do not include the cost of a cabin if reported earlier.

FOWN_ST In which state(s) was this land located?

FLEASE In 2016, did you, alone or with others, lease land in the United States

PRIMARILY for the purpose of fishing?

FLSE_ACRE How many acres did you lease?

FLSE_CLUB Were you part of a club or group, including a family group, leasing this land in 2016?

FLSE_CNUM How many others were members of this group or club, not counting

yourself?

FLSE_SHR What did you spend in 2016 for YOUR SHARE of the land which was leased PRIMARILY for fishing?

FLSE_ST In which state(s) was this land located?

EXEMPT_H Some hunters were exempt from BUYING a license in 2016 because of their

age, because they had a lifetime or free license, because they had a landowner exemption, or for some other reason. Were you exempt from buying a hunting license in any state in which you hunted in 2016?

EXEMPT_H_ST For which states?

BUY_H Did you buy a license to hunt in 2016? This could be a license that you

bought or was bought for you.


P a g e | 116

BUY_H_ST For which states?

NUM_H How many hunting licenses did you have for State Name in 2016?

COST_H Concerning your License_number license for State Name, how much did it cost?

OBT_H Did you actually obtain this license in 2016?

HUNT_H Was this license only for hunting, or to both hunt and fish?

DUCK_H Did you have a federal duck stamp in 2016?

FEES_H Some states charge special fees or require special permits, stamps, or tags

for certain types of hunting; a state waterfowl stamp or an elk permit, for example. Did you pay any such fees in 2016?

FEES_ST_H For which states?

FEES2H Concerning the special permits, stamps, and tags you bought in Hunting

State, in 2016, how much did they cost in total?

EXEMPT_F Some anglers were exempt from BUYING a license in 2016 because of their age, because they had a lifetime or free license, because they had a landowner exemption, or for some other reason. Were you exempt from buying a fishing license in any state in which you fished in 2016?

EXEMPT_F_ST For which states?

BUY_F Did you buy a license to fish in 2016? This could be a license that you

bought or was bought for you.

BUY_F_ST For which states?

NUM_F How many fishing licenses did you have for State Name in 2016?

FISH_F Concerning your License_number license for State Name, was this license only for fishing, or to both fish and hunt?

COST_F Concerning your License_number license for State Name, how much did it

cost?

OBT_F Did you actually obtain this license in 2016?


P a g e | 117

FEES_F Some states charge special fees or require special permits, stamps, or tags

for certain types of fishing; a trout stamp, for example. Did you pay any such fees in 2016?

FEES_ST_F For which states?

FEES2F Concerning the special permits, stamps, and tags you bought in Fishing State in 2016, how much did they cost in total?

H_PLNT During 2016, did you maintain any plantings, such as food or cover plants,

for the PRIMARY PURPOSE of hunting? Include areas in agricultural crops.

H_PLCST Approximately what were your costs for these plantings or crops during 2016?

FHINTRST Now I would like to ask you about observing and enjoying wildlife in ways

other than hunting and fishing. By wildlife, I mean birds, mammals, fish, insects, reptiles such as snakes and lizards, and amphibians such as frogs. Do not include farm animals or pets. From Date of last Sportsperson interview to December 31, 2016, did you take any special interest in wildlife around your home, other than simply noticing wildlife while doing other activities? By this I mean, did you closely observe, feed, or photograph wildlife within a one-mile radius of your home, OR did you maintain any natural areas or plantings around your home FOR WHICH BENEFIT TO WILDLIFE WAS AN IMPORTANT CONCERN? (Natural areas and plantings would include wooded lots, food and cover plants, etc.)

NCUINTRO Now I'm going to ask you some questions for a new time period, from Date

of last interview to December 31, 2016. "Wildlife" includes birds, mammals, fish, insects, reptiles such as snakes and lizards, and amphibians such as frogs.

FH_OBSRV From Date of Last Interview to December 31, 2016, did you take any trips or

outings in the United States of at least one mile from home for the PRIMARY PURPOSE of observing, photographing, or feeding wildlife? Do not include trips to zoos, circuses, aquariums, museums, or trips for hunting, fishing, or scouting for game.

NCU_ST In which state(s) did you take trips or outings to observe, photograph, or

feed wildlife?


P a g e | 118

USNCU From Date of last Wildlife Watcher interview to December 31, 2016, how

many days did you observe, photograph, or feed wildlife at least one mile from your home in the United States?

NCU_TOT How many trips lasting a single day or multiple days did you take in/to

Wildlife state from Date of last Wildlife Watcher interview to December 31, 2016 PRIMARILY to observe, photograph, or feed wildlife?

NCU_DAYS What is the total number of days you spent doing these activities in Wildlife

state?

TRIP On these trips/this trip in/to Wildlife state, did you... Observe wildlife? Photograph wildlife? Feed wildlife?

TRPDAY1 On these trips/this trip in/to Wildlife state, how many days did you observe

wildlife?

TRPDAY2 On these trips/this trip in/to Wildlife state, how many days did you photograph wildlife?

TRPDAY3 On these trips/this trip in/to Wildlife state, how many days did you feed

wildlife?

NCU_PRIV On your trips/trip in/to Wildlife state, did you visit any areas on privately-owned land?

NCU_PUB On your trips/trip in/to Wildlife state, did you visit any areas on land owned

by the local, State, or Federal Government?

BIRDS1 On your trips/trip in/to Wildlife state, did you observe birds?

BIRDS1A How many days did you observe birds?

BIRDS2 On your trips/trip in/to Wildlife state, did you photograph birds?

BIRDS3 On your trips/trip in/to Wildlife state, did you feed birds?

TYPBRD Did the birds you were observe in Wildlife state include...

ANIMLS On your trips/trip in/to Wildlife state, did you observe, photograph, or


P a g e | 119

feed...(read list of answer options).

NSINTRO Now I would like to ask you some questions about your expenses for all trips or outings that you took in the United States from Date of last Wildlife Watcher interview to December 31, 2016 for the PRIMARY purpose of observing, photographing, or feeding wildlife. If you paid for others or if someone else paid for you, INCLUDE ONLY YOUR SHARE OF THE EXPENSE.

NCUSHR_A On your trips/trip in/to Wildlife state, how much was spent for YOUR

SHARE of ... Food, drink, and refreshments?

NCUSHR_B Lodging at motels, cabins, lodges, campgrounds, etc.?

NCUSHR_C Public transportation by airplane?

NCUSHR_D Other public transportation, including trains, buses, and car rentals, etc.?

NCUSHR2_E The round-trip cost for transportation by private vehicle? Do not include

boating expenses.

NCUSHR2_F Guide fees, pack trip or package fees?

NCUSHR2_G Public land use or access fees? Do NOT include leases.

NCUSHR2_H Private land use or access fees? Do NOT include leases.

NCUSHR3_I Equipment rental such as boats, camping equipment, etc.?

NCUSHR3_J Boat fuel?

NCUSHR3_K Other boat costs (such as launching, mooring, storage, maintenance, pumpout fees, insurance)? Do not include land access fees already reported.

NCUSHR3_L Heating and cooking fuel?

NCUSHAR5_1 The total amount you spent on your trips/trip in/to Wildlife state was Total

wildlife trip expenses, not including airfare. How much of this was spent in your resident state of Resident state?

WILDLIFE Now I'm going to ask some questions about your experiences with wildlife


P a g e | 120

around your home. By "around your home," I mean the area within a one-mile radius of your home. From Date of last Wildlife Watcher interview to December 31, 2016, did you take SPECIAL INTEREST in wildlife around your home, other than simply noticing wildlife while doing other activities? By this I mean, did you closely observe wildlife or try to identify types of wildlife you did not know?

WILDDAYS How many days did you do this kind of observing of wildlife?

TYPWLD Which of the following kinds of wildlife did you observe?

TYPWLD1A How many days did you observe birds?

PHOTO Did you photograph any type of wildlife around your home?

PHOTDAY How many days?

FEEDBRD From Date of last Wildlife Watcher interview to December 31, 2016, did

you feed wild birds around your home?

FEEDFSH Did you feed any kind of fish or wildlife, other than birds, around your home? Please do not include animals you fed unintentionally.

PARKS From Date of last Wildlife Watcher interview to December 31, 2016, did

you visit any parks or natural areas within a one-mile radius of your home, for the purpose of observing, photographing, or feeding wildlife?

DYSPARK How many days did you visit these areas?

NCUEQP_A Now I would like to ask you about equipment and other items purchased

PRIMARILY for use in observing, photographing, or feeding fish or wildlife. Include only items that were purchased in the United States. Include both new items and items previously owned by others. As I read the following list, please tell me which items you purchased for yourself or were purchased for you. From Date of last Wildlife Watcher interview to December 31, 2016, did you PURCHASE or ACQUIRE - Binoculars, spotting scopes, etc.?


P a g e | 121

NCUEQP_B Cameras, special lenses, videocameras, or other photography equipment, including memory cards?

NCUEQP_C Film and photo processing?

NCUEQP2_D Commercially prepared and packaged wild bird food?

NCUEQP2_E Other bulk food used to feed wild birds?

NCUEQP2_F Food used to feed other wildlife?

NCUEQP2_G Nest boxes, bird houses, feeders, or baths?

NCUEQP2_H Any other purchases such as field guides, maps, etc.?

OTH_SPEC What was that purchase?

BNOCOST What was the total cost of the BINOCULARS, SPOTTING SCOPES, ETC.

purchased?

BNO_ST In which state(s) did you purchase this equipment?

PHOTOCST What was the total cost of the CAMERAS, SPECIAL LENSES, VIDEO CAMERAS, OR OTHER PHOTOGRAPHY EQUIPMENT, INCLUDING MEMORY CARDS purchased?

PHOTST In which state(s) did you purchase these items?

FILMCOST What was the total cost of the FILM AND PHOTO PROCESSING purchased?

FILMST In which state(s) did you purchase these items?

FOODCOST What was the total cost of the COMMERCIALLY PREPARED AND PACKAGED

WILD BIRD FOOD purchased?

FOODST In which state(s) did you purchase this bird food?

BULKCOST What was the total cost of the OTHER BULK FOOD USED TO FEED WILD BIRDS purchased?

BULKST In which state(s) did you purchase this other bird food?

OTHRCOST What was the total cost of the FOOD USED TO FEED OTHER WILDLIFE


P a g e | 122

purchased?

OTHRST In which state(s) did you purchase this food?

NESTCOST What was the total cost of the NEST BOXES, BIRD HOUSES, FEEDERS, OR BATHS purchased?

NESTST In which state(s) did you purchase this equipment?

SPECCOST What was the total cost of the Entry in OTH_SPEC purchased?

SPECST In which state(s) did you purchase this equipment?

MOREQP_A Show wildlife equipment reference card number

As I read the following list, please tell me if you purchased or acquired any other equipment primarily for use in observing, photographing, or feeding fish or wildlife from Date of last Wildlife Watcher interview to December 31, 2016. Did you PURCHASE or ACQUIRE - Tents, tarps?

MOREQP_B Frame packs, backpacking equipment?

MOREQP_C Other camping equipment?

MOREQP2_D Day packs, carrying cases, or special clothing (such as foul weather gear,

camouflage clothing, boots, etc.)?

MOREQP2_E Books, magazines, and DVD's specifically devoted to fish or wildlife?

MOREQP2_F Dues or contributions to national, state, or local conservation or wildlife-related organizations?

MOREQP2_G Any other purchases (such as blinds or GPS devices. Do not include boats,

cabins, or vehicles)?

OTHER_G What was that purchase?

TENTCOST How much did the TENTS AND/OR TARPS cost?


P a g e | 123

TENTST In which state(s) did you purchase this equipment?

PACKCOST How much did the FRAME PACKS AND/OR BACKPACKING EQUIPMENT cost?

PACKST In which state(s) did you purchase this equipment?

CAMP2CST How much did the OTHER CAMPING EQUIPMENT cost?

CMP2ST In which state(s) did you purchase this equipment?

DYPK_CST How much did the DAY PACKS, CARRYING CASES, OR SPECIAL CLOTHING cost?

DYPKST In which state(s) did you purchase this equipment?

MAG_COST How much did the BOOKS, MAGAZINES, OR DVD'S SPECIFICALLY DEVOTED

TO FISH OR WILDLIFE cost?

MAG_ST In which state(s) did you make this purchase?

DUE2COST What was the total cost of the DUES OR CONTRIBUTIONS TO CONSERVATION OR WILDLIFE-RELATED ORGANIZATIONS?

DUE2ST In which state(s) did you make this purchase?

OTH_GCST How much did the Entry in OTHER_G cost?

OTHGST In which state(s) did you purchase this equipment?

ANNINTRO In the remaining equipment questions, I will be referring to the ENTIRE

calendar year of 2016 - that is, from January 1 to December 31, 2016.

AnnEqp_A From this next list of equipment, please tell me which items you purchased for yourself or were purchased for you PRIMARILY for use in observing, photographing, or feeding fish or wildlife. During 2016, did you PURCHASE or ACQUIRE - An off-the-road vehicle (snowmobile, 4-wheeler, ATV, 4 by 4 vehicle, trail bike, dune buggy)?

AnnEqp_B A pickup, camper, van, travel or tent trailer, motor home, house trailer, or

recreational vehicle (RV)?


P a g e | 124

AnnEqp_C A boat, either motorized or not?

AnnEqp_D Boat accessories such as motor, trailer, or hitch?

AnnEqp_E A cabin?

AnnEqp_F Any other purchases?

AnnOth_F What was that purchase?

OTR_COST What was the amount paid for the OFF-THE-ROAD VEHICLE in 2016?


OTR_ST In which state(s) did you purchase this equipment?

TRAILCST What was the amount paid for the TRAVEL OR TENT TRAILER, MOTOR HOME, PICKUP, CAMPER, HOUSE TRAILER, RECREATIONAL VEHICLE (RV) OR VAN in 2016? Please only report the cost in 2016, not the total value of the item.

TRAILST In which state(s) did you purchase this equipment?

BOAT_CST What was the amount paid for the BOAT (EITHER MOTORIZED OR NOT) in

2016? Please only report the cost in 2016, not the total value of the item.

BOATSTN In which state(s) did you purchase this equipment?

ACCESCST What was the amount paid for the BOAT ACCESSORIES SUCH AS MOTOR, BOAT TRAILER/HITCH in 2016? Please only report the cost in 2016, not the total value of the item.

ACCEST In which state(s) did you purchase this equipment?

CABNCOST What was the amount paid for the CABIN in 2016? Please only report the

cost in 2016, not the total value of the item.

CABNST In which state(s) did you purchase the CABIN?

EQUIPCST What was the amount paid for the ^ANNOTH_F in 2016? Please only report the cost in 2016, not the total value of the item.

EQP_ST In which state(s) did you purchase this equipment?


P a g e | 125

AOWN Now I would like to ask you about any land that you owned or leased

because of your interest in observing, photographing, or feeding wildlife. In 2016, did you own land in the United States, in part or whole, PRIMARILY for the purpose of observing, photographing, or feeding wildlife?

AOWN_ACRE How many acres did you own?

AOWN_CLUB Were you part of a club or group, including a family group, owning this land

in 2016?

AOWN_CNUM How many others were members of this group or club, not counting yourself?

AOWN_SHR What did you spend in 2016 for YOUR SHARE of the land which was owned

PRIMARILY for observing, photographing, or feeding wildlife? Include mortgage, taxes, maintenance, and down payment cost if purchased in 2016. Do not include the cost of a cabin if reported earlier.

AOWN_ST In which state(s) was this land located?

ALEASE In 2016, did you lease land in the United States, alone or with others,

PRIMARILY for the purpose of observing, photographing, or feeding wildlife?

ALSE_ACRE How many acres did you lease?

ALSE_CLUB Were you part of a club or group, including a family group, leasing this land

in 2016?

ALSE_CNUM How many others were members of this group or club, not counting yourself?

ALSE_SHR What did you spend in 2016 for YOUR SHARE of the land which was leased

PRIMARILY for observing, photographing, or feeding wildlife?

ALSE_ST In which state(s) was this land located?

NATRAREA Some people maintain natural areas around their homes; that is, wooded lots, hedgerows, open fields, or other areas that provide a place for wildlife. During 2016, did you maintain any natural areas around your home for the PRIMARY PURPOSE of benefitting fish or wildlife? Include only areas 1/4 acre or more in size and do not include areas that are farmed.


P a g e | 126

ANNNUM1 How many acres of natural area did you keep or maintain for the benefit of

wildlife?

MAINPLNT During 2016, did you maintain in the area around your home any plantings, such as food or cover plants, for the PRIMARY PURPOSE of benefitting fish or wildlife? Include areas in agricultural crops.

ANPLCST Approximately what were your costs for these plantings or crops during

2016?

NCU_HNT Did you hunt game or other wildlife from Date of last Wildlife Watcher interview to December 31, 2016?

NCU_HNTST In which state(s)?

NCU_FISH Did you do any recreational fishing, including shellfishing, from Date of last

Wildlife Watcher interview to December 31, 2016?

NCU_FISHST In which state(s)?

Appendix C – Variance Items 2016 FHWAR Design and Methodology Technical Report

P a g e | 127

APPENDIX C. VARIANCE ITEMS

SUMMARY OF SCREENER VARIABLES INCLUDED ON ALL PUBLIC USE FILES AND VARIANCE ITEMS BY SAMPLE

Table C.1 details the screener variables included in the three public use files. Tables C.2-C.4 list,

by sample, the direct variance items that were calculated.

Table C.1 Screener Variables Included in All Three Public Use Files

Variable Description

CENDIV Census Division I_RESIDENT State (respondent) lived in at the start of Wave 1 interviewing RESSTATE State (respondent) lived in at the start of Wave 3 interviewing LETTER_A LETTER_B LETTER_C

Did you receive our letter with the enclosed reference aid? A,B,C in variable name refers to interview wave

MODE_A Mode of interview

MODE_B MODE_C

A,B,C in variable name refers to interview wave

GEMSAST Metropolitan Statistical Area (MSA) designator GEMSASZ Population size of Metropolitan Statistical Area GEUR Population density MARITAL Is (respondent) now – married, widowed, divorced, separated or

never married? SCHOOL What is the highest grade (or year) of regular school (respondent)

ever attended? HISPANIC Is (respondent) of Spanish, Hispanic, or Latino origin? RACE Did (respondent) specify a race (one of RACE1-RACE5 is filled)? SCRACE The variable is filled from answers in RACFE1-RACE5 RACE1 Did (respondent) specify their race as White? RACE2 Did (respondent) specify their race as Black or African American? RACE3 Did (respondent) specify their race as American Indian or Alaska

Native? RACE4 Did (respondent) specify their race as Asian? RACE5 Did (respondent) specify their race as Native Hawaiian or Other Pacific

Islander? HINCOME What was your total HOUSEHOLD income during 2015 before taxes

and other deductions?


P a g e | 128


Table C.2 Screener Variance Items

$137 Wildlife participant in 2016 (HUNT06, FISH06, INTEREST06) $2 Wildlife participant in 2015 (HUNT05, FISH05, INTEREST05) HUNT06 Did name hunt so far in 2016 FISH06 Did name fish so far in 2016 INTEREST06 Did name take any special interest in wildlife so far in 2016 HUNT05 Did name hunt in 2015 FISH05 Did name fish in 2015 INTEREST05 Did name take any special interest in wildlife in 2015

Table C.3 Sportsperson Variance Items


USHUNT Did you hunt in 2016 $3 How many hunting trips did you take (BGTRIPS, SMTRIPS, MBTRIPS,

OATRIPS) USDAYS_H How many days did you hunt in the US BGHUNT Did you hunt big game BG_TRIP How many trips did you take in reported state to hunt big game BG_DAYS How many days did you hunt big game in reported state SMHUNT Did you hunt small game SM_TRIP How many trips did you take in reported state to hunt small game SM_DAYS How many days did you hunt small game in reported state MBHUNT Did you hunt migratory birds MB_TRIP How many trips did you take in reported state to hunt migratory birds MB_DAYS How many days did you hunt migratory birds in reported state OAHUNT Did you hunt other animals OA_TRIP How many trips did you take in reported state to other animals OA_DAYS How many days did you hunt other animals in reported state RECFISH Did you recreationally fish in 2016 $4 How many fishing trips did you take (FWTRIPS, SWTRIPS) USDAYS_F How many days did you fish in the U.S. FSTATE Did you do any freshwater fishing in reported state FR_TRIP How many trips lasting a single day or more did you take to go

freshwater fishing FR_DAYS How many days did you fish in freshwater GSTATE Did you do any Great Lakes fishing in reported state GL_TRIP How many trips lasting a single day or more did you take to go Great

Lakes fishing GLSTDAYS How many days did you fish in Great Lakes

37 If the Variable column contains a “$” then this indicates the question uses multiple variables to reach the final answer/total; the variables used can be found at the end of the description. The same goes for Tables C4 and C4.


P a g e | 129

Variable Description SSTATE Did you do any saltwater fishing in reported state SALTTRIP How many trips lasting a single day or more did you take to go

saltwater fishing SALTDAYS How many days did you fish in saltwater $5 Sportsman (USHUNT, USFISH) $6 How many days did you fish in other states (FISHST, I_RESIDENT,

STDAYSF, USDAYS_F) $7 How many days did you fish in state of residence (FISHST, I_RESIDENT,

STDAYSF, USDAYS_F) $8 Did you freshwater fish in other states, excluding Great Lakes (FRSTE,

I_RESIDENT) $9 Did you saltwater fish in state of residences (SLTSTE, I_RESIDENT) $10 Did you saltwater fish in other states (SLSTE, I_RESIDENT) FRTYP Freshwater fished except Great Lakes – black bass DAYS_FR Days fishing - black bass FRTYP Freshwater fished except Great Lakes – panfish DAYS_FR Days fishing - panfish DAYS_FR Days freshwater fished - catfish and bullheads FRTYP Freshwater fished except Great Lakes – trout DAYS_FR Days fishing - trout GLTYP Great Lakes fished – walleye, sauger GLDAYS Days fishing - walleye, sauger GLTYP Great Lakes fished – perch GLDAYS Days fishing - perch GLDAYS Days Great Lakes fished - salmon SLTYP Saltwater fished – striped bass SALTDAY Days fishing saltwater - striped bass $11 Did you hunt in other states (HUNTST, I_RESIDENT) $12 Did you hunt big game in state of residence (HUNTST, I_RESIDENT,

BGHNT) HUNT Big game hunted – deer BGDIFDAY Days hunting – deer HUNT Big game hunted - wild turkey BGDIFDAY Days hunting - wild turkey HUNT Small game hunted – rabbit SMDIFDAY Days hunting – rabbit HUNT Small game hunted – quail SMDIFDAY Days hunting – quail HUNT Small game hunted – squirrel SMDIFDAY Days hunting – squirrel HUNT Small game hunted – pheasant SMDIFDAY Days hunting – pheasant


P a g e | 130

Variable Description HUNT Migratory bird hunted – geese MBDIFDAY Days hunting - geese HUNT Migratory bird hunted – duck MBDIFDAY Days hunting - duck $13 Amount spent on fishing or hunting equipment (RIFLECOST,

GUNSCOST, MUZZCST, PISTLCST, BOWSCOST, SCOPECST, DECOYCST, AMMCOST, HANDCST, DOGSCOST, OTHCOST, RODSCOST, LINECOST, LURECOST, HOOKCOST, TACKCOST, CREELCST, BAITCOST, FINDCOST, ICECOST, OTHERCST)

$14 Amount spent on auxiliary equipment - fishing or hunting (CAMPCOST, BINOCOST, GEARCOST, TAXICOST, FHCOST)

BOOKCOST Amount spent on books, magazines, and DVD's for fishing or hunting DUECOST Total cost of dues or contributions $15 Amount spent on heating and cooking fuel – fishing (OFSHAR11,

SLSHAR11, GLSHAR11) $16 Amount spent on fishing equipment (RODSCOST, LINECOST,

LURECOST, HOOKCOST, TACKCOST, CREELCST, BAITCOST, FINDCOST, ICECOST, OTHERCST)

RODSCOST Total cost of rods, reels, poles, and rod making components purchased HOOKCOST Total cost of hooks, sinkers, swivels, etc. purchased LURECOST What was the total cost of the artificial lures, flies, baits TACKCOST Total cost of tackle boxes CREELCST Total cost of creels, stingers, fish bags, landing nets, and gaff hooks BAITCOST Total cost of minnow traps, seines, and bait containers $17 Amount spent on fishing license (COSTF, OBTF, FISHF, USHUNT,

COSTH, OBTH, HUNTH) $18 Amount spent on camping equipment – fishing (CAMPCOST,

CAMPUSE) OFSHAR_A OF - Amount spent on food, drink, and refreshments OFSHAR_B OF - Amount spent on lodging $19 Amount spent on transportation – Freshwater (OFSHAR3, OFSHAR4,

OFSHAR5) OFSHAR2_G OF - Amount spent on public land use or access fees OFSHAR4_N OF - Amount spent on boat launching fees GLSHAR_A Amount spent on food, drink, and refreshment - Great Lakes $20 Amount spent on fishing – Saltwater (SLSHAR1-SLSHAR15, RODSCOST,

LINECOST, LURECOST, HOOKCOST, TACKCOST, CREELCST, BAITCOST, FINDCOST, LICECOST, OTHERCST, RODSUSE, LINEUSE, LUREUSE, HOOKUSE, TACKUSE, CREELUSE, BAITUSE, FINDUSE, ICEUSE, OTHERUSE, CAMPCOST, BINOCOST, GEARCOST, TAXICOST, FHCOST, CAMPUSE, BINOUSE, GEARUSE, TAXIUSE, FHUSE, CAMPFISH, BINOFISH, GEARFISH, TAXIFISH, FHFISH, BASSCOST, BOATCOST,


P a g e | 131

Variable Description CANOECST, MOTORCST, VANCOST, CABINCST, OFF_COST, EQPCST, BASSUSE, BOATUSE, CANOEUSE, MOTORUSE, VANUSE, CABINUSE, OFFUSE, OTHG_USE, BASSFISH, BOATFISH, CANOEFISH, MOTORFSH, VANFISH, CABINFSH, OFF_FISH, OTHG_FISH)

$21 Amount spent on fishing trip related costs – Saltwater (SLSHAR1-SLSHAR15)

SLSHAR_E Amount spent for transportation by private vehicle Saltwater SLSHAR_F Amount spent on guide fees, pack trip or package fees SLSHAR2_K SW - Amount spent on heating and cooking fuel SLSHAR2_L SW - Amount spent on equipment rental - boats, fishing or camping

equipment, etc. SLTOTAMOUNT Total Amount spent on saltwater fishing trips $22 Amount spent on hunting trip related costs (BGSHAR1-BGSHAR13,

SMSHAR1-SMSHAR13, MBSHAR1-MBSHAR13, OASHAR1-OASHAR13) $23 Amount spent on lodging – hunting (BGSHAR2, SMSHAR2, MBSHAR2,

OASHAR2) $24 Amount spent on other trip costs – hunting (BGSHAR6-BGSHAR13,

SMSHAR6-SMSHAR13, MBSHAR6-MBSHAR13, OASHAR6-OASHAR13) $25 Amount spent on hunting equipment (RIFLECST, GUNSCOST,

MUZZCST, PISTLCST, BOWSCOST, SCOPECST, DECOYCST, AMMCOST, HANDCST, DOGSCOST, OTHCOST)

RIFLECST Amount spent on rifles GUNCOST Amount spent on shotguns BOWSCOST Amount spent on bows, arrows or other archery equipment SCOPECST Amount Spent on telescopic sights $26 Amount spent on special equipment for hunting (BASSCOST,

BOATCOST, CANOECST, MOTORCST, VANCOST, CABINCST, OFF_COST, EQPCST, BASSUSE, BOATUSE, CANOEUSE, MOTORUSE, VANUSE, CABINUSE, OFFUSE, OTHG_USE)

$27 Amount spent on hunting magazines (BOOKCOST, BOOKUSE) $28 Amount spent on hunting dues and contributions (DUECOST,

DUESUSE) $29 Amount spent on leasing and ownership of land for hunting

(HOWN_SHR, HLSE_SHR) $30 Amount spent on hunting licenses (COSTH, OBTH) FEES2H Amount spent on special permits, stamps, and tags $31 Amount spent on hunting transportation (BGSHAR3-BGSHAR5,

SMSHAR3-SMSHAR5, MBSHAR3-MBSHAR5, OASHAR3-OASHAR5) AMMCOST What was the total cost of ammunition $32 Amount spent on big game hunting (BGSHAR1-BGSHAR13, RIFLECST,

GUNSCOST, MUZZCST, PISTLCST, BOWSCOST, SCOPECST, RIFLEUSE, GUNSUSE, MUZZUSE, PISTLUSE, BOWSUSE, SCOPEUSE, DECOYUSE,


P a g e | 132

Variable Description AMMUSE, HANDUSE, DOGSUSE, OTHUSE, CAMPCOST, BINOCOST, GEARCOST, TAXICOST, FHCOST, CAMPUSE, BINOUSE, GEARUSE, TAXIUSE, FHUSE, CAMPHNT, BINOHNT, GEARHNT, TAXIHNT, FHHNT, BASSCOST, BOATCOAST, CANOECST, MOTORCST, VANCOST, CABINCST, OFF_COST, EQPCST, BASSUSE, BOATUSE, CANOEUSE, MOTORUSE, VANUSE, CABINUSE, OFFUSE, OTHG_USE, BASSHNT, BOATHNT, CANOEHNT, MOTORHNT, VANHNT, CABINHNT, OFF_HNT, OTHG_HNT)

$33 BG - Amount. spent on food and lodging (BGSHAR1, BGSHAR2) BGSHAR_E BG - Amount spent on round trip cost for transportation by private

vehicle $34 BG - Amount. spent on special hunting clothes (GEARCOST, GEARUSE,

GEARHNT) $35 BG - Amount. spent on taxidermy costs (TAXICOST, TAXIUSE, TAXIHNT) $36 Amount spent on small game hunting (SMSHAR1-SMSHAR13,

RIFLECST, GUNSCOST, MUZZCST, PISTLCST, BOWSCOST, SCOPECST, DECOYCST, AMMCOST, HANDCST, DOGSCOST, OTHCOST, RIFLEUSE, GUNSUSE, MUZZUSE, PISTLUSE, BOWSUSE, SCOPEUSE, DECOYUSE, AMMUSE, HANDUSE, DOGSUSE, OTHUSE, CAMPCOST, BINOCOST, GEARCOST, TAXICOST, FHCOST, CAMPUSE, BINOUSE, GEARUSE, TAXIUSE, FHUSE, CAMPHNT, BINOHNT, GEARHNT, TAXIHNT, FHHNT, BASSCOST, BOATCOST, CANOECST, MOTORCST, VANCOST, CABINCST, OFF_COST, EQPCST, BASSUSE, BOATUSE, CANOEUSE, MOTORUSE, VANUSE, CABINUSE, OFFUSE, OTHG_USE, BASSHNT, BOATHNT, CANOEHNT, MOTORHNT, VANHNT, CABINHNT, OFF_HNT, OTHG_HNT)

SMSHAR_A SG - Amount spent of food, drink, and refreshments $37 SG - Amount. spent on transportation (SMSHAR3, SMSHAR4,

SMSHAR5) $38 Amount spent on migratory bird hunting (MBSHAR1-MBSHAR13,

RIFLECST, GUNSCOST, MUZZCST, PISTLCST, BOWSCOST, SCOPECST, DECOYCST, AMMCOST, HANDCST, DOGSCOST, OTHCOST, RIFLEUSE, GUNSUSE, MUZZUSE, PISTLUSE, BOWSUSE, SCOPEUSE, DECOYUSE, AMMUSE, HANDUSE, DOGSUSE, OTHUSE, CAMPCOST, BINOCOST, GEARCOST, TAXICOST, FHCOST, CAMPUSE, BINOUSE, GEARUSE, TAXIUSE, FHUSE, CAMPHNT, BINOHNT, GEARHNT, TAXIHNT, FHHNT, BASSCOST, BOATCOST, CANOECST, MOTORCST, VANCOST, CABINCST, OFF_COST, EQPCST, BASSUSE, BOATUSE, CANOEUSE, MOTORUSE, VANUSE, CABINUSE, OFFUSE, OTHG_USE, BASSHNT, BOATHNT, CANOEHNT, MOTORHNT, VANHNT, CABINHNT, OFF_HNT, OTHG_HNT)

MBSHAR_A MB – Amount spent for food and drink GEARCOST Total cost of special fishing or hunting clothing


P a g e | 133

Variable Description $39 Amount spent on special fishing or hunting equipment (BASSCOST,

BOATCOST, CANOECST, MOTORCST, VANCOST, CABINCST, OFF_COST, EQPCST)

$40 Did you use a boat to fish (BOATGL, BOATFR, BOATSL) BOATCOST Amount spent on other types of motor boats BUY_F Did you buy a license to fish in 2016 EXEMPT_F Were you exempt from buying a fishing license BUY_H Did you buy a license to hunt EXEMPT_H Were you exempt from buying a hunting license DAYPOND How many days did you fish in ponds or lakes or reservoirs FR_RIVER Did you fish in reported state(s) in rivers or streams RIVERDAY How many days did you fish in rivers or streams WHCHGL Fished - Lake Erie DAYS_GL Days fished - Lake Erie WHCHGL Fished - Lake Michigan DAYS_GL Days fished - Lake Michigan $41 Hunted on public land (BGPUBLIC, SMPUBLIC, MBPUBLIC, OAPUBLIC) $42 Days of hunting on public land (BGDYSPUB, SMDYSPUB, MBDYSPUB,

OADYSPUB, STDAYSH) BG_PRVT Did you do any big game hunting in reported state on privately-owned

land BGDYPRV How many days did you hunt big game in reported state on privately-

owned land BGPUBLIC Did you do any big game hunting in reported state on public land BGDYSPUB How many days did you hunt big game in reported state on public

land SM_PRVT Did you do any small game hunting in reported state on privately-

owned land SMDYPRV How many days did you hunt small game in reported state on

privately-owned land SMPUBLIC Did you do any small game hunting in reported state on public land SMDYSPUB How many days did you hunt small game in reported on public land MBPUBLIC Did you do any migratory bird hunting in reported state on public land MBDYPUB How many days did you hunt migratory birds in reported state on

public land MB_PRVT Did you do any migratory bird hunting in reported state on privately-

owned land MBDYPRV How many days did you hunt migratory birds in reported state on

privately-owned land ICEFISH Did you ice fish DAYS_IF How many days did you ice fish FLYFISH Did you fly fish


P a g e | 134

Variable Description DAYSFLY How many days did you fly fish FLEASE Did you lease land in the U.S. primarily for fishing FH_EQP_A Did you purchase or acquire - camping equipment FH_EQP_B Did you purchase or acquire - binoculars, field glasses, telescopes

Table C.4 Wildlife-Watching Variance Items


FH_OBSRV Nonresidential participant $43 Did you participate in a state other than state of residence (NCU_ST,

I_RESIDENT, I_WAVE3) AOWN Did you own land primarily for observing, photo, or feeding wildlife NCU_TOT How many trips did you take in reported state(s) USNCU How many days did you observe, photo or feed at least 1 mile from

home AOWN_SHR What did you spend on this owned land TRPDAY1 Days observing wildlife TRPDAY2 Days photographing wildlife TRPDAY3 Days feeding wildlife $44 Did you observe, photo, or feed birds (BIRDS1, BIRDS2, BIRDS3) ANIMLS Did you observe, photo, or feed fish ANIMLS Did you observe, photo, or feed land mammals ANIMLS Did you observe, photo, or feed marine mammals NCUSHR_A Amount spent on - food, drinks, and refreshments NCUSHR_B Amount spent on - lodging $45 Amount spent on public transportation (NCUSHR3, NCUSHR4) NCUSHR2_E Amount spent on - round trip transportation by private vehicle NCUSHR2_F Amount spent on - guide fees, pack trip, or package fees NCUSHR2_G Amount spent on - public land use or access fees NCUSHR2_H Amount spent on - private land use or access fees NCUSHR3_I Amount spent on - equipment rental: boats, camping, etc. $46 Residential user (USRESIDE) $47 Wildlife participant (FH_OBSRV, I_WAVE3, USRESIDE) WILDLIFE Did you take special interest in wildlife around your home WILDDAYS How many days did you do this kind of observing TYPWLD Did you observe birds TYPWLD Did you observe mammals PHOTO Did you photograph any type of wildlife around your home PHOTDAY How many days did you photograph $48 Feed wildlife around home (FEEDBRD, FEEDFSH) PARKS Did you visit any parks within one mile of your home to observe,

photo of feed DYSPARK How many days did you visit these parks


P a g e | 135

Variable Description BNOCOST Total cost of binoculars purchased FILMCOST Total cost of film purchased PHOTOCST Total cost of cameras, etc. purchased DYPK_CST Total cost of carrying cases, special clothing purchased FOODCOST Total cost of wild bird food purchased BULKCOST Total cost of bulk food purchased NESTCOST Total cost of nest boxes, etc. purchased $49 Amount spent on wildlife-watching equipment (BINOCOST,

PHOTOCST, FILMCOST, FOODCOST, BULKCOST, OTHRCOST, NESTCOST, DYPK_CST, SPECCOST)

MAG_COST Total cost of books, magazines purchased DUE2COST Total cost of dues purchased OTH_GCST Total cost of other items purchased TENTCOST Total cost of tents, tarps purchased PACKCOST Total cost of frame packs purchased CAMP2CST Total cost of other camping equip purchased $50 Amount spent on auxiliary equipment (TENTCOST, PACKCOST,

CAMP2CST, OTH_GCST) OTR_COST Amount paid for - off-the-road vehicle TRAILCST Amount paid for - trailer, etc. EQUIPCST Amount paid for - other equipment $51 Did you maintain any natural areas or plantings around your home for

the primary purpose of benefitting wildlife (NATRAREA, MAINPLNT) $52 Total trip expenditures (NCUSHR1-NCUSHR12) OBSERVE Did you observe wildlife - outside of resident state PHOTOGRAPH Did you photograph wildlife - outside of resident state FEED Did you feed wildlife - outside of resident state

Appendix D – Comparability Across 2016 FHWAR Design and Methodology Technical Report FHWAR Surveys

P a g e | 136

APPENDIX D. COMPARABILITY ACROSS FHWAR SURVEYS

Table D.1 Anglers and Hunters by Census Division: 1991, 1996, 2001, 2006, 2011, and 2016

(U.S. population 16 years and older. Numbers in thousands)

Area and sportsperson

1991 1996 2001 2006 2011 2016

Number Percent Number Percent Number Percent Number Percent Number Percent Number Percent

United States

Total population 189,964 100 201,472 100 212,298 100 229,245 100 239,313 100 254,686 100

Sportspersons 39,979 21 39,694 20 37,805 18 33,916 15 37,397 16 39,553 16

Anglers 35,578 19 35,246 17 34,067 16 29,952 13 33,112 14 35,754 14

Hunters 14,063 7 13,975 7 13,034 6 12,510 5 13,674 6 11,453 4

New England


Sportspersons 1,658 16 1,673 16 1,504 14 1,353 12 1,441 12 1,485 12

Anglers 1,545 15 1,520 15 1,402 13 1,246 11 1,355 12 1,333 11

Hunters 444 4 465 5 386 4 374 3 420 4 297 2

Middle Atlantic


Sportspersons 4,508 15 4,192 14 3,810 13 3,214 10 3,966 12 3,793 11

Anglers 3,871 13 3,627 12 3,250 11 2,550 8 3,496 11 3,471 10

Hunters 1,746 6 1,453 5 1,633 5 1,520 5 1,558 5 884 3

East North Central


Sportspersons 7,202 22 6,912 21 6,400 19 5,975 17 6,766 19 7,097 19

Anglers 6,264 19 6,006 18 5,655 17 5,190 15 5,861 16 6,336 17

Hunters 2,789 9 2,712 8 2,421 7 2,376 7 2,688 7 2,737 7

West North Central


Sportspersons 4,143 31 3,977 29 4,239 29 3,836 25 3,980 25 3,487 21

Anglers 3,647 27 3,416 25 3,836 27 3,284 21 3,591 23 3,042 18

Hunters 1,709 13 1,917 14 1,710 12 1,779 12 1,661 10 1,364 8

South Atlantic


Sportspersons 6,996 21 7,282 20 6,957 18 6,633 15 6,749 15 8,181 16

Anglers 6,441 19 6,636 18 6,451 16 6,116 14 6,163 13 7,394 15

Hunters 2,083 6 2,050 6 1,875 5 1,884 4 1,870 4 1,716 3

East South Central


Sportspersons 2,984 26 2,907 23 2,865 22 2,689 20 3,010 21 3,386 23

Anglers 2,635 23 2,514 20 2,543 20 2,436 18 2,444 17 3,061 20

Hunters 1,279 11 1,301 10 1,164 9 1,101 8 1,531 11 *1,256 *8

West South Central


Sportspersons 5,125 26 5,093 23 4,924 21 4,499 18 4,855 18 5,694 19

Anglers 4,592 23 4,616 21 4,375 19 3,952 16 4,298 16 5,206 17

Hunters 1,843 9 1,812 8 1,988 9 1,810 7 1,909 7 1,556 5

Mountain


Sportspersons 2,488 25 2,761 23 2,757 21 2,372 15 2,976 17 2,941 16

Anglers 2,079 21 2,411 20 2,443 18 2,084 13 2,586 15 2,687 15

Hunters 1,069 11 1,061 9 1,020 8 868 6 1,043 6 946 5

Pacific


Sportspersons 4,875 17 4,897 15 4,349 13 3,345 9 3,654 10 3,489 8

Anglers 4,505 15 4,501 14 4,111 12 3,094 8 3,319 9 3,224 8

Hunters 1,101 4 1,203 4 837 2 798 2 996 3 697 2


P a g e | 137

Table D.2 Wildlife-Watching Participants by Census Division: 1991, 1996, 2001, 2006, 2011, and 2016


Area and wildlife watcher

1991 1996 2001 2006 2011 2016

Number Percent Number Percent Number Percent Number Percent Number Percent Number Percent

United States

Total population

Wildlife watchers

Away from home

Around the home

New England

Total population

Wildlife watchers

Away from home

Around the home

Middle Atlantic

Total population

Wildlife watchers

Away from home

Around the home

East North Central

Total population

Wildlife watchers

Away from home

Around the home

West North Central

Total population

Wildlife watchers

Away from home

Around the home

South Atlantic

Total population

Wildlife watchers

Away from home

Around the home

East South Central

Total population

Wildlife watchers

Away from home

Around the home

West South Central

Total population

Wildlife watchers

Away from home

Around the home

Mountain

Total population

Wildlife watchers

Away from home

Around the home

Pacific

Total population

Wildlife watchers

Away from home

Around the home

189,964

76,111

29,999

73,904

10,180

4,598

1,856

4,544

29,216

10,556

4,166

10,282

32,188

14,511

5,572

14,175

13,504

6,924

2,654

6,722

33,682

13,047

4,450

12,813

11,667

4,864

1,592

4,765

19,926

7,035

2,459

6,817

10,092

4,437

2,215

4,145

29,508

10,139

5,035

9,641

100

40

16

39

100

45

18

45

100

36

14

35

100

45

17

44

100

51

20

50

100

39

13

38

100

42

14

41

100

35

12

34

100

44

22

41

100

34

17

33

201,472

62,868

23,652

60,751

10,306

3,710

1,443

3,586

29,371

8,185

2,960

8,023

33,121

11,731

4,501

11,297

13,875

5,089

1,927

4,900

36,776

11,252

3,992

10,964

12,459

3,904

1,118

3,795

21,811

5,933

2,096

5,773

11,966

4,099

1,967

3,855

31,787

8,966

3,648

8,558

100

31

12

30

100

36

14

35

100

28

10

27

100

35

14

34

100

37

14

35

100

31

11

30

100

31

9

30

100

27

10

26

100

34

16

32

100

28

11

27

212,298

66,105

21,823

62,928

10,575

3,875

1,155

3,765

29,806

8,740

2,849

8,452

34,082

11,631

3,571

11,196

14,430

6,206

2,059

5,938

39,286

11,395

3,469

10,911

12,976

4,514

1,086

4,390

23,337

5,747

1,822

5,490

13,308

4,619

2,019

4,282

34,498

9,377

3,793

8,504

100

31

10

30

100

37

11

36

100

29

10

28

100

34

10

33

100

43

14

41

100

29

9

28

100

35

8

34

100

25

8

24

100

35

15

32

100

27

11

25

229,245

71,132

22,977

67,756

11,233

4,489

1,340

4,310

31,518

8,723

2,729

8,451

35,609

12,215

3,792

11,845

15,458

6,741

2,163

6,447

43,965

12,862

3,208

12,432

13,722

4,931

1,758

4,683

25,407

6,764

2,127

6,319

15,651

4,968

2,004

4,605

36,681

9,439

3,856

8,664

100

31

10

30

100

40

12

38

100

28

9

27

100

34

11

33

100

44

14

42

100

29

7

28

100

36

13

34

100

27

8

25

100

32

13

29

100

26

11

24

239,313

71,776

22,496

68,598

11,593

3,954

1,187

3,858

32,392

9,118

2,561

8,744

36,199

12,840

3,168

12,492

15,860

5,479

1,783

5,201

46,417

13,315

4,393

12,767

14,206

4,663

1,456

4,394

27,195

7,164

1,728

7,087

17,013

5,189

2,230

4,716

38,438

10,054

3,990

9,337

100

30

9

29

100

34

10

33

100

28

8

27

100

35

9

35

100

35

11

33

100

29

9

28

100

33

10

31

100

26

6

26

100

30

13

28

100

26

10

24

254,686

86,042

23,720

81,128

12,018

4,430

1,499

4,336

33,368

12,170

3,688

11,838

36,893

13,348

2,847

12,808

16,502

5,322

1,590

5,249

50,611

17,832

5,530

16,502

14,968

5,062

*498

4,907

30,094

8,173

1,541

7,763

18,364

6,257

3,119

4,883

41,869

13,448

3,408

12,842

100

34

9

32

100

37

12

36

100

36

11

35

100

36

8

35

100

32

10

32

100

35

11

33

100

34

*3

33

100

27

5

26

100

34

17

27

100

32

8

31


P a g e | 138

Table D.3 Comparison of Major Findings of the National Surveys: 1955 to 1985


Sportspersons 1955 1960 1965 1970 1975 1980 1985

Total sportspersons 24,917 30,435 32,881 36,277 45,773 46,966 49,827 Anglers 20,813 25,323 28,348 33,158 41,299 41,873 45,345 Freshwater 18,420 21,677 23,962 29,363 36,599 35,782 39,122 Saltwater 4,557 6,292 8,305 9,460 13,738 11,972 12,893

Hunters 11,784 14,637 13,583 14,336 17,094 16,758 16,340 Small game 9,822 12,105 10,576 11,671 14,182 12,496 11,130 Big game 4,414 6,277 6,566 7,774 11,037 11,047 12,576 Waterfowl 1,986 1,955 1,650 2,894 4,284 3,177 3,201

1 Expenditures $11,401,464 $13,948,974 $14,991,502 $19,618,548 $33,398,677 $34,517,421 $42,058,860 Anglers $7,655,522 $9,743,971 $9,952,411 $13,699,311 $23,498,506 $23,387,469 $28,585,686 Freshwater $5,700,187 $7,476,454 $7,231,851 $10,315,966 $17,333,212 $16,663,239 $18,942,060 Saltwater $1,955,336 $2,267,512 $2,720,574 $3,383,345 $6,165,294 $5,581,976 $7,191,387

Hunters $3,745,942 $4,204,997 $3,814,303 $5,919,236 $9,900,171 $10,812,058 $10,256,668 Small game $1,975,707 $2,629,360 $2,093,137 $2,612,390 $4,525,942 $3,335,852 $2,342,860 Big game $1,295,357 $1,251,800 $1,424,711 $2,631,532 $4,238,341 $5,638,395 $5,345,606 Waterfowl $474,878 $323,840 $296,452 $675,315 $1,135,889 $766,033 $783,315

Days 566,870 658,308 708,578 909,876 1,459,551 1,300,983 1,415,379 Fishing 397,447 465,769 522,759 706,187 1,058,075 952,420 1,065,986 Freshwater 338,826 385,167 426,922 592,494 890,576 788,392 895,027 Saltwater 58,621 80,602 95,837 113,694 167,499 164,040 171,055

Hunting 169,423 192,539 185,819 203,689 401,476 348,543 350,393 Small game 118,630 138,192 128,448 124,041 269,653 225,793 214,544 Big game 30,834 39,190 43,845 54,536 100,600 117,406 135,447 Waterfowl 19,959 15,158 13,526 25,113 31,223 26,179 25,933 1 In 1985 U.S. dollars Note: Methodological differences described in Chapter 2 make the estimates in this table not comparable with the estimates in Tables D.1 and D.2.


P a g e | 139

Table D.4 Anglers and Hunters by Census Division: 1955 to 1985


Year Population Sportsperson, fished or hunted Anglers Hunters

Number Percent Number Percent Number Percent Number Percent

United States 1955 1960 1965 1970 1975 1980 1985 New England 1955 1960 1965 1970 1975 1980 1985 Middle Atlantic 1955 1960 1965 1970 1975 1980 1985 East North Central 1955 1960 1965 1970 1975 1980 1985 West North Central 1955 1960 1965 1970 1975 1980 1985 South Atlantic 1955 1960 1965 1970 1975 1980 1985 East South Central 1955 1960 1965 1970 1975 1980 1985

118,366

131,226

141,928

155,230

171,860

184,691

195,659

118,366

7,919

8,349

9,256

8,652

9,910

10,205

10,554

24,869

26,493

27,346

28,244

30,449

30,256

31,099

25,733

26,833

28,124

31,550

32,796

33,526

33,747

9,201

10,149

11,681

12,904

13,564

13,826

14,137

14,336

17,798

20,593

23,539

27,127

30,512

33,636

7,959

9,277

9,652

9,862

10,798

11,771

12,364

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

24,917

30,435

32,881

36,277

45,773

46,966

49,827

24,917

1,224

1,368

1,650

1,579

2,004

1,974

2,058

3,539

3,432

3,602

4,539

5,919

5,181

5,565

5,489

6,316

6,214

7,284

9,049

8,725

8,973

2,913

3,383

3,678

4,000

4,524

4,770

5,140

3,223

4,423

5,626

5,461

7,110

7,769

8,721

1,963

2,778

2,587

2,660

3,007

3,614

3,671

21.1

23.2

23.2

23.4

26.6

25.4

25.5

21.1

15.4

16.4

17.8

18.3

20.2

19.3

19.5

14.2

13.0

13.2

16.1

19.4

17.1

17.9

21.3

32.5

22.1

23.1

27.6

26.0

26.6

31.7

33.3

31.5

31.0

33.3

34.5

36.4

22.5

24.9

27.3

23.2

26.2

25.5

25.9

24.7

29.9

26.8

27.0

27.8

30.7

29.7

20,813

25,323

28,348

33,158

41,299

41,873

45,345

20,813

1,002

1,205

1,488

1,430

1,861

1,788

1,914

2,811

2,569

2,760

4,504

5,097

4,332

4,820

4,583

5,317

5,336

6,699

8,181

7,891

8,270

2,346

2,855

3,226

3,579

4,089

4,220

4,681

2,805

3,695

5,054

5,129

6,479

7,086

8,056

1,665

2,207

2,201

2,464

2,689

3,173

3,308

17.6

19.3

20.0

21.4

24.0

22.7

23.2

17.6

12.7

14.4

16.0

16.5

18.8

17.5

18.1

11.3

9.7

10.1

14.4

16.7

14.3

15.5

17.8

19.8

19.0

21.2

24.9

23.5

24.5

25.5

28.1

27.6

27.7

30.1

30.5

33.1

19.6

20.8

24.5

21.8

23.9

23.2

24.0

20.9

23.8

22.8

25.0

24.9

27.0

26.8

11,784

14,637

13,585

14,336

17,094

16,758

16,340

11,784

589

517

583

582

566

572

552

1,608

1,723

1,631

1,731

2,096

2,001

1,972

2,538

2,985

2,563

2,812

3,392

2,955

2,814

1,534

1,709

1,620

1,783

1,863

1,965

1,971

1,449

2,045

1,900

1,904

2,494

2,444

2,467

989

1,510

1,294

1,162

1,355

1,567

1,441

10.0

11.2

9.6

9.2

9.9

9.1

8.4

10.0

7.4

6.2

6.3

6.7

5.7

5.6

5.2

6.5

6.5

6.0

6.1

6.9

6.6

6.3

9.9

11.1

9.1

8.9

10.3

8.8

8.3

16.7

16.8

13.9

13.8

13.7

14.2

13.9

10.1

11.5

9.2

8.1

9.2

8.0

7.3

12.4

16.3

13.4

11.8

12.5

13.3

11.7


Population Sportsperson, fished or hunted Anglers Hunters Year

Number Percent Number Percent Number Percent Number Percent

West South Central

1955 10,250 100 2,560 25.0 2,237 21.8 1,165 11.4

1960 11,837 100 3,666 31.0 3,133 26.5 1,750 14.8

1965 12,724 100 3,713 29.2 3,278 25.8 1,571 12.3

1970 14,624 100 4,380 30.0 4,006 27.4 1,918 13.1

1975 16,628 100 5,781 34.8 5,267 31.7 2,563 15.4

1980 19,136 100 5,862 30.6 5,136 26.8 2,456 12.8

1985 21,184 100 6,418 30.3 5,704 26.9 2,572 12.1

Mountain 1955 4,529 100 1,369 30.2 1,112 24.6 796 17.6

1960 5,222 100 1,646 31.5 1,372 26.3 1,120 21.4

1965 5,029 100 1,565 31.1 1,261 25.1 988 19.6

1970 5,656 100 2,044 36.1 1,769 31.3 980 17.3

1975 7,576 100 2,570 33.9 2,252 29.7 1,159 15.3

1980 9,160 100 2,903 31.7 2,500 27.3 1,268 13.8

1985 10,215 100 3,128 30.6 2,765 27.1 1,241 12.1

Pacific 1955 13,570 100 2,637 19.4 2,252 16.6 1,116 8.2

1960 15,268 100 3,422 22.4 2,971 19.5 1,279 8.4

1965 17,523 100 4,246 24.2 3,744 21.4 1,433 8.2

1970 20,199 100 4,332 21.4 4,030 20.0 1,466 7.3

1975 23,012 100 5,811 25.2 5,386 23.4 1,607 7.0

1980 26,299 100 6,168 23.5 5,747 21.9 1,531 5.0

1985 38,725 100 6,154 21.4 5,829 20.3 1,310 4.6

Note: Methodological differences described in Chapter 2 make the estimates in this table not comparable with the estimates in Tables D.1 and D.2.

P a g e | 140

Appendix E – Outcome Codes 2016 FHWAR Design and Methodology Technical Report

P a g e | 141

APPENDIX E. OUTCOME CODES

Table E.1 defines the CATI outcome codes and Table E.2 defines the CAPI outcome codes. These

codes were used for both the household-level data collections (Waves 1 and 2) and the person-

level data collection (Wave 3).

Table E.1 CATI Outcome Codes

Outcome Code Description

001

002

006

007

020

024

025

026

179

181

182

183

188

193

194

195

Fully Complete

Complete Screener, sufficient partial

Complete/Partial

Complete Screener, no one in sample

Sample Unit Ineligible - out of scope

Unconverted language problem

Unconverted hearing barrier

In scope but data unavailable

Hostile breakoff

Refusal

Hard refusal

Exceeded unproductive call max

Insufficient partial - callback

Privacy detector

Never contacted - confirmed number

Never contacted - unconfirmed number

Appendix E – Outcome Codes 2016 FHWAR Design and Methodology Technical Report

P a g e | 142

Table E.2 CAPI Outcome Codes

Outcome Code Description

201 Complete Screener/Detail

204 Complete Screener, no one in sample

206 Complete Screener, detail undone – transmitted

208 Complete Screener, detail sufficient partial

216 No one home

217 Temporarily absent

218 Refused

219 Type A - Other

225 Temp Occupied with Usual Residence Elsewhere (URE)

226 Vacant

228 Unfit/to be demolished

230 Converted to temp business

231 Unoccupied tent/trailer site

233 Type B - Other

234 Type B - Institutionalized

240 Demolished

241 House/trailer moved

243 Converted to business

244 Merged

245 Condemned

248 Type C – Other

250 Type C, deceased

251 Type C, respondent moved out of country

260 Address Unknown - no further research possible

national survey of fishing, hunting, and wildlife ... · 12 years and older personal inter-view, 12...

Documents